New Zealand Shipping v A M Satterthwaite (The Eurymedon)

On 8 December 1995 the vessel Starsin sailed from the Far East bound for ports in Western Europe. Among other cargoes the vessel carried a number of parcels of timber and plywood the condition of which deteriorated progressively and seriously during the voyage because they had been negligently stowed before the voyage began. These parcels were carried pursuant to contracts of carriage contained in or evidenced by a series of transferable bills of lading of which the respondents, on different dates, became the holders. It is convenient to describe the respondents as "the cargo owners". They made claims for damage to the cargo against the appellants, the owners and demise charterers of the vessel, between whom it is unnecessary to distinguish and to whom I shall refer as "the shipowner". The vessel was, at the time of this voyage, on time-charter by the shipowner to Continental Pacific Shipping ("CPS"), which at the time operated a liner service between the Far East and Western Europe, but which has since become insolvent. This appeal by the shipowner and the cross-appeal by the cargo owners raise questions which, compendiously expressed, are whether the shipowner is liable to the cargo owners under these bill of lading contracts and, if not, whether it is liable to any of the cargo owners to any (and if so, what) extent in tort.

The facts are well summarised in the judgments of Colman J sitting in the Commercial Court ([2000] 1 Lloyd's Rep 85) and the Court of Appeal ( [2001] 1 Lloyd's Rep 437 , [2001] EWCA Civ 56 ) and those summaries need not be repeated. It is enough to note that the various parcels of timber and plywood were shipped on board the vessel at 3 ports in Malaysia for carriage to Antwerp and Avonmouth under 17 bills of lading of which each of the 4 cargo owners (Makros Hout BV, Homburg Houtimport BV, Fetim BV and Hunter Timber Ltd) became the holders of 2 or more. Save in certain respects, specifically mentioned below (paragraph 38), it is unnecessary to distinguish between the various cargo owners. There were some differences between the various bills, but these differences are not material to any issue arising on the appeals. It is convenient to confine my description to a single bill, one of the Makros Hout bills, which I shall call "the bill" and treat as generally representative of all the bills sued upon.

The bill was in familiar form. It had 2 sides, a face and a reverse. On its face, the bill contained in its top left-hand corner a box with the heading "Shipper" into which the name and address of the shipper under the bill was typed. Below that box was the heading "Consigneee", and here was typed "To order of Hongkong Bank Malaysia Berhad". Below that was a box ("Notify address") into which the name and address of Makros Hout were typed. Below that was the heading "Vessel" and in that box was typed "MV Starsin V.CP144", the latter abbreviation standing (it appears) for "Voyage Continental Pacific No 144". The ports of loading and final destination were typed into boxes provided to contain that information. On the top right-hand corner of the face of the bill was the bold heading "Liner Bill of Lading" with a notation making abbreviated reference to the ports of loading and discharge. Below that, much more prominent than any other entry on the face of the bill, was a logo and the printed words "Continental Pacific Shipping". Below the entries so far described, also on the face of the bill, particulars were typed in, as supplied by the shipper, of the marks, numbers of packages, description, weight and measurement of the goods. Below these particulars a box was provided for details of freight and charges to be entered, and "Freight prepaid" was typed in. To the right of this box was a printed statement to this effect:

Below this statement were boxes providing for details (which were inserted) of the place where freight was to be paid, the number of original bills, the description of the particular bill and the place and date of issue of the bill. Lastly, in the bottom right-hand corner of the bill was a box with the printed heading "Signature". In that box there was typed "As Agent for Continental Pacific Shipping (The Carrier)", below which words was a rubber stamp ("United Pansar Sdn Bhd"), that being a company which acted as port agent for CPS at Kuching (the port of loading under the bill), and across the box were what appear to be 2 manuscript signatures.

On the reverse of the bill was a prominent printed heading "Company's Standard Conditions", below which 35 conditions were set out, in 2 dense columns, in print which was very small but just legible by a painstaking and persistent reader. Particular reliance was placed on conditions 1 (the definitions clause), 33 (the identity of carrier clause) and 35 (the demise clause). In some of the bills certain of the conditions became garbled in the printing, giving rise to obvious errors of spelling, punctuation, grammar and meaning, but I do not think it is in doubt that these 3 conditions should be treated as reading, in all the bills, as follows:

The first and most crucial issue between the parties on these appeals is whether the contracts to carry these various parcels of cargo were made by or on behalf of the shipowner, as the cargo owners contend, or by or on behalf of CPS, the charterers of the vessel, as the shipowner contends. Put another way, the question is whether these were shipowner's bills or charterer's bills. Colman J held that they were charterer's bills and that CPS was the contractual carrier: [2001] 1 Lloyd's Rep 85. On appeal, the Court of Appeal was divided: [2001] 1 Lloyd's Rep 437 . Rix LJ agreed with the trial judge. But Sir Andrew Morritt V-C and Chadwick LJ held that they were shipowner's bills and that the shipowner was the contractual carrier, entitling the cargo owners to sue in contract for any recoverable loss they had suffered.

In submitting that these bill of lading contracts should be understood as having been made with CPS, Mr Gee QC for the shipowner relied above all on the form of signature on the face of the bill: the port agents signed expressly "As Agent for Continental Pacific Shipping (The Carrier)". Thus the contract was made on behalf of CPS as the party undertaking to carry the goods. The "Carrier" was defined in condition 1(c) to mean "the party on whose behalf this Bill of Lading has been signed", namely CPS. If there was inconsistency between that unambiguous statement of CPS's role just above the signature and general descriptions elsewhere in the text, the former should be regarded as the "dominating factor" and as of "preponderant importance": Universal Steam Navigation Company Ltd v James McKelvie and Company [1923] AC 492, per Lord Shaw of Dunfermline at p.499, per Lord Sumner at p.500. Words which the parties have themselves chosen and written into the contract should have greater effect than printed standard terms: Glynn v Margetson & Co [1893] AC 351 .

In submitting that these were shipowner's and not charterer's bills, as the Court of Appeal majority had held, Mr Milligan QC for the cargo owners relied on a number of features of the bill clearly indicating that it was intended to take effect as a contract entered into by the shipowner. Provision was made on the face for signature by the master of the vessel, a strong although not conclusive pointer towards an owner's bill: Wehner v Dene Steam Shipping Co [1905] 2 KB 92 at 98; The Rewia [1991] 2 Lloyd's Rep 325 at 333. The attestation clause made reference to all the stipulations of the bill "on both pages, whether written, printed, stamped or otherwise incorporated". Conditions 33 and 35 in particular, but also other conditions in the bill such as conditions 16 and 17, were intended to make plain that the contract of carriage was made with the owner (or demise charterer) of the vessel, with whom sole liability should rest to the exclusion of any liability falling on the charterer or any agent. Thus Mr Milligan submitted that on a proper construction of the whole of the bill the shipowner should be held to be the contracting party, perhaps by regarding the shipowner as the disclosed but unnamed principal of CPS, perhaps by construing the contract as made with the shipowner as well as CPS, perhaps because the description of CPS as carrier was unauthorised by CPS and so ineffective. There is in my opinion no evidence to show that CPS contracted as agent for the shipowner as disclosed but unnamed principal, and the terms of the signature are inconsistent with that suggestion. There is, again, nothing to suggest dual liability in CPS and the shipowner; the standard conditions are expressed in the singular throughout, with no provision that the singular shall include the plural. There is no evidence to show that the port agent lacked authority to sign in the terms it did. But the problem of construction remains.

When construing a commercial document in the ordinary way the task of the court is to ascertain and give effect to the intentions of the contracting parties. Here, the task is to ascertain who, on one side, the contracting party was. But a similar approach is appropriate. Mr Milligan urged that the House should not seize on a single canon of construction and give it effect to the exclusion of all others. I am sure that warning is salutary. But there are a number of rules, some of very long standing, which give valuable guidance.

First is the rule to which Lord Halsbury alluded in Glynn v Margetson & Co [1893] AC 351 at 359, "that a business sense will be given to business documents". The business sense is that which businessmen, in the course of their ordinary dealings, would give the document. It is likely to be a reasonably straightforward sense since, as Lord Mansfield famously observed ( Hamilton v Mendes (1761) 2 Burr 1198 at 1214, 97 ER 787 at 795),

Secondly, it is common sense that greater weight should attach to terms which the particular contracting parties have chosen to include in the contract than to pre-printed terms probably devised to cover very many situations to which the particular contracting parties have never addressed their minds. It is unnecessary to quote the classical statement of this rule by Lord Ellenborough in Robertson v French (1803) 4 East 130 at 136; 102 ER 779 at 782, cited with approval by Lord Halsbury in Glynn v Margetson [1893] AC 351 at 358 and by Scrutton LJ in In re an Arbitration between L Sutro & Co and Heilbut, Symons & Co [1917] 2 KB 348 at 361-2.

Thirdly, it has long been recognised by very distinguished commercial judges that to seek perfect consistency and economy of draftsmanship in a complex form of contract which has evolved over many years is to pursue a chimera: see, for example, Simond v Boydell (1779) 1 Dougl 268; 99 ER 175; James Nelson & Sons Ltd v Nelson Line (Liverpool) Ltd [1908] AC 16 at 20-21; Hillas & Co Ltd v Arcos Ltd (1932) 43 Ll. L. Rep 359 at 367; Chandris v Isbrandtsen-Moller Co Inc [1951] 1 KB 240 at 245. The court must of course construe the whole instrument before it in its factual context, and cannot ignore the terms of the contract. But it must seek to give effect to the contract as intended, so as not to frustrate the reasonable expectations of businessmen. If an obviously inappropriate form is used, its language must be adapted to apply to the particular case: The Okehampton [1913] P 173 at 180, per Hamilton LJ.

Fourthly,

It is plain, for reasons which Mr Milligan gave, that the bill was drafted to express or evidence a contract between the shipper (and any transferee of the bill) and the owner of the vessel. Conditions 33 and 35 so state. The provision for signature by the master of the vessel so indicates. But a very cursory glance at the face of the bill is enough to show that the master has not signed the bill. It has instead been signed by agents for CPS which is described as "The Carrier". I question whether anyone engaged in maritime trade could doubt the meaning of "carrier", a term of old and familiar meaning, but any such doubt would be quickly resolved by resort to the first condition overleaf in which the term is defined to mean the party on whose behalf the bill of lading has been signed, that is, the party contracting to carry the goods.

I can well understand that a shipper or transferee of a bill of lading would recognise the need to consult the detailed conditions on the reverse of the bill in any one of numerous contingencies which might arise and for which those conditions make provision. He would appreciate that the rights and obligations of the parties under the contract are regulated by those detailed conditions. But I have great difficulty in accepting that a shipper or transferee of a bill of lading would expect to have to resort to the detailed conditions on the reverse of the bill (and to persevere in trying to read the conditions until reaching conditions 33 and 35) in order to discover who he was contracting with. And I have even greater difficulty in accepting that he would expect to do so when the bill of lading contains, on its face, an apparently clear and unambiguous statement of who the carrier is.

I am fortified in adopting this view of market practice by noting that, although it was not adopted by Moore-Bick J in The Flecha [1999] 1 Lloyd's Rep 612, it was adopted by Rix J in The Hector [1998] 2 Lloyd's Rep 287 (in which there was, as here, an identity of carrier clause but not, as here, a demise clause), it was adopted by Colman J in the present case ([2000] 1 Lloyd's Rep 85 at 93) and it was adopted by Rix LJ in his persuasive dissent on this point in the present case ( [2001] 1 Lloyd's Rep 437 at 451). I am further fortified in taking this view of market practice by noting its adoption (since 1994) in the ICC Uniform Customs and Practice for Documentary Credits . This now provides, in Article 23(a) that

I would note, lastly on this point, that the decision of the Court of Appeal majority has not earned the approval of some academic commentators expert in this field. Professor Debattista (Lloyd's List, 21 February 2001, p.5 "Is the end in sight for chartering demise clauses?") opined that the dissenting judgment of Rix LJ and its robust market sense were to be preferred to the opinion of the Court of Appeal majority. The authors of "Contracts for the Carriage of Goods by Land Sea and Air" (LLP, ed Yates, service issue 21, 31 December 2001, p.1-380, para 1.6.4.2.25.1) similarly suggest that

I agree with these opinions and would hold, in agreement with Colman J and Rix LJ, and for essentially the reasons which they gave, that the bill contained or evidenced a contract of carriage made with CPS as carrier.

The second issue arising in these appeals is whether, assuming CPS to be the contractual carrier under these contracts of carriage, the terms of the contracts are effective to protect the shipowner against liability to the cargo owners. The answer must turn on the terms and applicability of clause 5 of the printed conditions on the reverse of the bills, which is headed "IMMUNITIES". There are some differences between the fine detail of this provision in the Homburg bills as compared with the other bills, but it is not suggested that these are of legal significance. I shall continue to treat the bill (as defined above) as representative.

As it appears in the bill, clause 5 is a very lengthy provision and most of it is irrelevant for present purposes. So far as relevant the clause reads as follows:

It is plain that the printing of this clause leaves something to be desired. For example, some punctuation is missing and the word "over" in the last sentence should obviously read "were". It is also plain, and common ground, that there is some omission at the point where I have inserted the asterisk in square brackets. Colman J resolved this problem by interpolating "(who)": [2000] 1 Lloyd's Rep 85 at 98. Rix LJ made the same interpolation, as did Sir Andrew Morritt V-C, although the latter set out the clause somewhat differently: [2001] 1 Lloyd's Rep 437 at 460, 476. Argument was addressed to the House, not addressed to the courts below, on the correctness of this interpolation. While the answer to this problem is not decisive in these appeals, it is as well to resolve how the clause should be read before construing it.

While acknowledging that there was no justification for inserting "who" other than a need to correct an obvious grammatical solecism, Mr Milligan urged that, since it was not the function of the courts to make or re-write the parties' contracts, the interpolation to be made should be the least intrusive reasonably possible to make sense of the clause. Mr Gee contended (a) that it was clear what words have been omitted, and (b) that the omission was explained by the phenomenon, technically known as homoeoteleuton, where one sentence contains a word which closed the preceding sentence and the transcriber's eye has wandered from one to the other, leading to the entire omission of the whole passage lying between them. As to (a), Mr Gee contended that the missing words were "acting as aforesaid and for the purpose of all the foregoing provisions of this clause the Carrier … ." These are the words to be found in the Conline bill of lading form, on which clause 5 as quoted above has been closely modelled, although with some additions, deletions and changes of language. It is a form of bill in very wide use, and was the subject of judicial consideration in A M Satterthwaite & Co Ltd v New Zealand Shipping Co Ltd (The Eurymedon) [1975] AC 154 at 165 and Port Jackson Stevedoring Pty Ltd v Salmond and Spraggon (Australia) Pty Ltd (The New York Star) [1981] 1 WLR 138 . After the square brackets the clause again picks up the language of the Conline bill. As to (b), Mr Gee submitted that the transcriber's eye had wandered from "Carrier" immediately before the square brackets to the same word at the end of the suggested interpolation.

I take it to be clear in principle that the court should not interpolate words into a written instrument, of whatever nature, unless it is clear both that words have been omitted and what those omitted words were: see Re Hargraves' Trusts, Leach v Leach [1937] 2 All ER 545 at 547-548, see per Farwell J; In re Murray, decd. Martins Bank Ltd v Dill [1955] Ch 69 at 79-80, per Evershed MR; In re Neeld , decd. Carpenter v Inigo-Jones [1960] Ch 455 at 464-465, per Cross J; [1962] Ch 643 at 677-678, per Upjohn LJ; In re Riley's Will Trusts. Riley v Riley [1962] 1 WLR 344 at 348-349, per Buckley J. In the present case there is agreed to be an omission. It is also plain, in my opinion, for the reasons which Mr Gee gave, what words were omitted and how they came to be omitted. I would accordingly construe the clause as if the words "acting as aforesaid and for the purpose of all the foregoing provisions of this clause the Carrier" appeared in place of the square brackets I have inserted.

It is now necessary to construe the clause to decide the extent (if any) of the protection it affords the shipowner. For this purpose the relevant part of the clause (as quoted above) must be construed as a whole, notwithstanding the convenience of breaking it down into its constituent numbered parts. Part (1) on its face confers wide-ranging immunity on any servant, agent or independent contractor of the carrier, in the course of his employment by the carrier, against any liability to the shipper resulting from (among other things) negligent damage to the goods. It is not suggested that the cargo owners are in any different position from the shippers. Colman J interpreted this provision as a covenant not to sue, enforceable by injunction, such as was considered in Nippon Yusen Kaisha v International Import and Export Co Ltd (The Elbe Maru) [1978] 1 Lloyd's Rep 206: [2000] 1 Lloyd's Rep 85 at 99-100. All three members of the Court of Appeal agreed with him; [2001] 1 Lloyd's Rep 437 at p 461, para 116; p 471, para 169; p 476, para 201. Given this unanimity of opinion, one is reluctant to disagree. But it is in my judgment impossible to spell a covenant not to sue out of the language of this clause, as the Court of Appeal found it to be in Gore v Van Der Lann [1967] 2 QB 31. The language construed in The Elbe Maru (above) was strikingly different.

Part (2) has two obvious purposes. The first is to extend to every servant, agent or independent contractor of the carrier acting as such every right and every exemption, defence and immunity available to the carrier. The second is to establish that the carrier is for purposes of all the foregoing provisions of the clause acting on behalf of all such servants, agents and independent contractors. The intention, plainly, is to fill the lacuna exposed in Adler v Dickson [1955] 1 QB 158 and to achieve the contractual solution based on agency which was presciently described by Lord Reid in Midland Silicones Ltd v Scruttons Ltd [1962] AC 446 at 474 and which was held to be effective in The Eurymedon [1975] AC 154 and The New York Star [1981] 1 WLR 138 . It is unnecessary to repeat, but it is helpful to bear in mind, the very clear account of the historical background to Himalaya clauses provided by Lord Goff of Chieveley, giving the judgment of the Privy Council, in The Mahkutai [1996] AC 650 at pp 658-665.

Part (3) again has two purposes. The first is to underline that such servants, agents and independent contractors are to be treated as parties to the contract. The second is to make plain that such servants, agents and independent contractors are only parties to the contract to a limited extent ("to this extent"): that is, while "acting as aforesaid" and for the purpose of asserting the rights, exemptions, defences and immunities already referred to.

Part (4) is an addition to the standard wording of the Conline bill. But it appears to be a straightforward indemnity provision, and it gives rise to no problem in these appeals.

A number of points were raised on the construction of this clause. The first was whether, assuming that CPS was the contractual carrier, the shipowner was an independent contractor. The judge held that it was ([2000] 1 Lloyd's Rep 85 at 99):

Rix and Chadwick LJJ and Sir Andrew Morritt V-C agreed: [2001] 1 Lloyd's Rep 437 at 461, para 113; p.471, para 166; pp.475-476, paras 198-201. 1 also agree. The judge, as I respectfully think, correctly analysed the role of an independent contractor and correctly held that, on the present facts, the shipowner fell within it.

The second question is whether the protection available to servants, agents and independent contractors under part (1) was by virtue of part (2) limited to the protection (to use a compendious term) available to the carrier under part (2). The judge held that the protection was so limited: [2000] 1 Lloyd's Rep 85 at 99-100. All three members of the Court of Appeal agreed: [2001] 1 Lloyd's Rep 437 at pp.461-2, paras 114-117; p.471, para 171; p.476, para 201. There are to my mind two insuperable objections to this construction. First, part (1) is expressed in unqualified (and, if it is relevant, emphatic) language: "in any circumstances whatsoever be under any liability whatsoever". Secondly, part (2) is prefaced by the words "without prejudice to the generality of the provisions in this Bill of Lading". This is the classical formula used by a draftsman to make plain that what follows is not to be understood as restricting the effect of some other provision or provisions. So, since part (1) is a provision in the bill, part (2) is not to be understood as restricting it. That seems to me a compelling reason for not understanding part (1) in any sense more restricted than its ordinary meaning would suggest. I would not regard it as an objection to this construction that part (2) appeared to be unnecessary, given the incidence of repetition and surplusage in instruments of this kind and the number of different jurisdictions in which the clause might fall to be construed. Some weight was attached in the courts below to the words "to this extent" in part (3), but I have indicated my understanding of these words in paragraph 26 above and I do not think they help the cargo owners.

Account must however be taken of the Hague Rules, which were expressly stipulated to be the basis of this contract of carriage by clause 2 on the reverse of the bill. The Hague Rules impose certain duties on the carrier, notably the duties in article III rules 1 and 2:

Article IV provides various grounds of exoneration and provides a provisional financial limitation applicable to claims for loss or damage to goods. Article III rule 8 provides

The cargo owners rely on this last rule to support their contention that part (1) must be understood to limit the protection of servants, agents and independent contractors of the carrier to the protection available to the carrier, since otherwise the exemption would fall foul of this rule. The counter-argument was that "carrier" is defined in article I(a) to include "the owner or the charterer who enters into a contract of carriage with a shipper" and that the party which had entered into a contract of carriage with the shipper on the present facts was CPS, not the shipowner. Therefore, it was said, there was no reason why the liability of the shipowner should not be excluded altogether. Until a late stage I was inclined to accept this counter-argument as correct, but I am ultimately persuaded that it is not, for a reason which did not, I think, loom large in the courts below or in argument before the House.

The Himalaya clause itself, and the undoubted artificiality of the reasoning relied on to uphold it in The Eurymedon [1975] AC 154 and The New York Star [1981] 1 WLR 138 , were a deft and commercially-inspired response to technical English rules of contract, particularly those governing privity and consideration: per Lord Goff in The Pioneer Container [1994] 2 AC 324 at 335 and The Mahkutai [1996] AC 650 at 664-665. In both The Eurymedon and The New York Star the problem arose because there was no contract between the cargo owner and the stevedore made in the ordinary way by a mutual exchange of promises: the stevedore had never promised to do anything. The problem was overcome by holding, in effect, that a bilateral contract came into existence upon the performance by the stevedore of stevedoring services, the stevedore thereupon becoming entitled to the benefit of the Himalaya clause exemption provided by the carrier acting as agent of the stevedore. In this situation the question whether the stevedore had itself become party to a contract of carriage could scarcely arise, since on no showing could it be said to have done so. The present case however is factually different, because the act performed to bring any contract into existence between the shipowner and the cargo owners is the carrying of the goods. The question is whether that factual difference gives rise to a legal difference, whether (in short) the resulting contract is properly to be regarded for Hague Rules purposes as a contract of carriage and the shipowner as entering into it with a shipper. I have not found these to be easy questions, but I conclude that to answer them negatively would be to elevate form over substance and to invest what is essentially a legal device with a wholly disproportionate legal significance. If the act performance of which brings a contract into existence between the shipowner and the cargo owners is the carrying of the cargo owners' goods it would seem to me anomalous to give the shipowner the benefit of clause 5 but take no account of article III rule 8 of the Hague Rules which were incorporated into the contract by clause 2 (where they were described as the "BASIS OF CONTRACT"). Thus the shipowner is not protected by an exemption provision invalidated by article III rule 8. Since the object of joining third parties to the contract was only to give them the benefit of the contractual exemption, that object was (in the case of the shipowner) frustrated. But since the object of joining third parties (here the shipowner) to the contract was only to give the shipowner the benefit of the contractual exemption, the shipowner did not become subject to the positive obligations laid on the carrier by article III rules 1 and 2 of the Hague Rules, and this was not suggested.

I would accordingly resolve this second issue in favour of the cargo owners, reaching the same destination as the Court of Appeal but by a different route. This makes it necessary to consider the third major issue in the case, which is whether the cargo owners (other than Makros Hout) have a tort claim maintainable against the shipowner.

In Leigh and Sillavan Ltd v Aliakmon Shipping Co Ltd (The Aliakmon) [1986] AC 785 at 809, Lord Brandon of Oakbrook, in an opinion with which all members of the House agreed, said:

The issue between the present parties concerned not the correctness of this principle, but its application to the facts of the present case, which were very economically summarised by Rix LJ ( [2001] 1 Lloyd's Rep 437 at 453, para 78 as follows:

Makros Hout had already obtained title to their goods before damage to those goods occurred. Thus they were owners when the damage occurred and their cause of action became complete. It follows that, the second issue having been answered in favour of the cargo owners, the shipowner has no defence to a claim in tort by Makros Hout. Another cargo owner, Hunter Timber, was unable to show when it acquired title to the goods and so was unable to show that it was the owner of the goods consigned to it when its cause of action arose. For the purpose of the ensuing discussion the claims of these two cargo owners must be excluded.

The Court of Appeal was unanimous in rejecting the judge's conclusion that because the damage was progressive each exacerbation of pre-existing damage gave rise to a new cause of action exercisable by the owner of the goods at that time, thus permitting an allocation of causes of action to different shippers and consignees and an apportionment of the resulting damage. Rix LJ ( [2001] 1 Lloyd's Rep 437 at 457, para 96) said

At that time neither Fetim nor Homburg were cargo owners. Rix LJ regarded Darley Main Colliery Co v Mitchell (1886) 11 App Cas 127 as applying to a special and exceptional situation: pp 457-459, paras 99-105. Chadwick LJ reached the same conclusion (p 464, para 132), as did Sir Andrew Morritt V-C (p.476, para 204). I would for my part also reach that conclusion, for all the reasons given by Rix LJ in paragraphs 77-107 of his judgment, at pages 453-459, which I am content to adopt as my own.

I would therefore dismiss the cargo owners' cross-appeal. But I would allow the shipowner's appeal, save in the case of Makros Hout, the judgment in whose favour must stand. I would invite the parties to make written submissions on costs in the House and in the courts below.

In 1995, while the Starsin was on time charter to Continental Pacific Shipping Limited ("CPS"), bills of lading on CPS' liner form were issued for the carriage of 17 parcels of timber and plywood. The carriage was between ports in Malaysia and Antwerp/Avonmouth. The cargoes were damaged. Receivers sued the owners for breach of contract, or alternatively in tort in the event that the bills of lading were charterers' bills.

On this appeal the central issues are:

The issue was whether the bills of lading were charterers' or owners' bills. The terms of each of the 17 bills reveal inconsistent provisions. CPS were the charterers. The signature box on the face or front of a specimen bill of lading prominently carried a signature "As Agent for Continental Pacific Shipping (The Carrier)." The face of the bill of lading contained essential commercial provisions such as the identity of the shippers, the name of the vessel and a description of the cargo, as well as a reference to the contract of carriage, the latter being set out in the box commencing with the word "Shipped." Clause 1(c) is consistent with what appears on the face of the bill of lading: it provides that the "carrier" is "the party on whose behalf this bill of lading has been signed." So far the document is in harmony. But tucked away in barely legible tiny print on the back of the bill of lading are two clauses which contradict the contractual position revealed by the face of the bill. Clause 33 provides that the contract evidenced by the bill of lading was "between the merchant and the owner of the vessel named herein (or substitute)." Clause 35, a demise clause, provides that the bill of lading shall only take effect as a contract of carriage "with the owners or demise charterers."

How is the problem to be addressed? For my part there is only one principled answer. It must be approached objectively in the way in which a reasonable person, versed in the shipping trade, would read the bill. The reasonable expectations of such a person must be decisive. In my view he would give greater weight to words specially chosen, such as the words which appear above the signature, rather than standard form printed conditions. Moreover, I have no doubt that in any event he would, as between provisions on the face of the bill and those on the reverse side of the bill, give predominant effect to those on the face of the bill. Given the speed at which international trade is transacted, there is little time for examining the impact of barely legible printed conditions at the time of the issue of the bill of lading. In order to find out who the carrier is it makes business common sense for a shipper to turn to the face of the bill, and in particular to the signature box, rather than clauses at the bottom of column two of the reverse side of the bill.

Taking advantage of their knowledge of the way in which the market works two commercial judges - Colman J and Rix LJ in the Court of Appeal - adopted the mercantile view. The majority in the Court of Appeal - Morritt V-C and Chadwick LJ - in effect gave preponderant effect to the boilerplate clauses on the back of the bill. In my view it would have an adverse effect on international trade if the latter approach prevails. Professor Debattista ( Is the end in sight for chartering demise clauses?, Lloyds List, Wednesday 21 February 2001, 5) rightly warned that the effect of the judgment of the majority would be to create traps for the unwary: see also the criticism of the majority judgments by Professor Gaskell and others in Contracts for the Carriage of Goods by Land, Sea and Air , LLP, ed. Yates, as updated by service issue No. 21, dated 31 December 2001, para 1.6.4.2.25.1, and by Dr Girvin and Professor Bennett in English Maritime Law 2000 , [2002] LMCLQ, at 84-87. As Rix LJ observed, commercial certainty and indeed honesty is promoted by giving greater effect to the front of the bill of lading.

This conclusion is reinforced by the ICC Uniform Customs and Practice for Documentary Credits: 1993 revision in force as of January 1, 1994. Article 23(a) reads as follows:

It follows that in my judgment the ruling of the majority on the identity of the carrier point cannot stand. I would also hold that the ex tempore judgment in Fetim BV v Oceanspeed Shipping Limited (The Flecha) [1999] 1 Lloyd's Rep 612 was wrong.

Counsel for the cargo owners raised an alternative argument. He argued that words like "The Carrier," "For the Carrier," and "As Carrier," can be treated as adding the personal liability of CPS rather than excluding the liability of the owners. This is a question of interpretation. Counsel for the owners showed convincingly that the bill of lading contemplated a single carrier. It is only necessary to mention specifically that the completed signature box, as well as the definition clause, points to a single carrier. I would reject this alternative argument.

I conclude that CPS was the sole carrier under the bill of lading. The owners were not parties to the contract of carriage and are not liable under the bill of lading contracts. Their potential liability in tort must now be considered.

The issue whether the owners are protected from liability in tort by the Himalaya clause now arises.

The owners rely in particular on that part of clause 5 of the bills of lading which Lord Bingham has recited and labelled part (1). The owners contend that part (1) confers on them a general exemption from liability.

Before this question of construction can be considered a preliminary issue must be examined. As Lord Bingham has pointed out by reference to the Himalaya clause contained in the Conline bill of lading form some words have been left out of clause 5: see the clauses in New Zealand Shipping Co Ltd v A M Satterthwaite & Co Ltd (The Eurymedon) [1975] AC 154 and Port Jackson Stevedoring Pty Ltd v Salmond & Spraggon (Australia) Pty Ltd (The New York Star) [1981] 1 WLR 138 . The deletion was plainly a mistake. In these circumstances the court should, in order to give effect to the reasonable expectations of the parties, fill the gap by inserting what had been omitted. What falls to be construed is the clause so reconstructed.

The result is that it cannot be argued that part (1) is flawed by reason of a lack of agency or authority. Accordingly part (1) must take effect in accordance with the terms of clause 5 as reconstructed.

For my part the language of part (1), contextually considered, is capable of one interpretation only. The shipowners are clearly "independent contractors." It is not capable of being read as a mere covenant not to sue. In categorical language part (1) confers on the shipowners a general exemption from liability.

The various attempts by the cargo owners to argue that part (1) is "not clear enough" are on examination at variance with the benign advance heralded by The Eurymedon , carried forward by The New York Star [1981] 1 WLR 138 and reaffirmed in The Mahkutai [1996] AC 650 . In The New York Star Lord Wilberforce commended a wide interpretation of the reasoning in The Eurymedon . He said at 144:

In my view the arguments of the cargo owners are of the very type which Lord Wilberforce warned against. I would respectfully also echo an extra-judicial statement by of Lord Goff of Chieveley in Commercial Contracts and the Commercial Court (1984) LMCLQ 382, 391:

That brings me to the question of the impact of the incorporated Hague Rules on clause 5. Clause 2 (described as the "Basis of Contract") incorporated the Hague Rules. This neutral fact tells us nothing about the impact of the Hague Rules on the Himalaya clause. That depends on the proper construction of the relevant Hague Rules. Of course, the ship owner will not be entitled to rely on the Himalaya clause if it offends the Hague Rules. The question is whether the exemption in favour of the ship owners in fact is in conflict with the Hague Rules.

I turn to the critical provisions of the Hague Rules. Article III rule 8 of the Hague Rules. This provision reads as follows:

One cannot, however, construe article III rule 8 in isolation. Article III rules 1 and 2 of the Hague Rules are relevant. They read as follows:

I have noted the reasoning based on the words in para (3) that the independent contractor "shall to this extent be deemed to be [a party] to the contract contained in or evidenced by this Bill of Lading". For my part this places a weight on those words which they will not bear.

For my part this result is in no way anomalous. It is loyal to the rationale of the advance in the rationality of English law achieved in The Eurymedon and The New York Star . It results in a readily predictable scheme, viz all claims in contract and tort have to be channelled to the charterers. That gives effect to what the parties intended to achieve. It has the merit of being a just decision achieved without in any way straining the Hague Rules. I would hold that the exemption contained in part (1) of clause 5 protects the owners against any liability in tort.

Regretfully, and in the spirit of accepting that one must not be too confident that one is right, I consider it appropriate to record my disagreement with the majority on this issue.

This issue falls away. Given that it has been fully argued I accept, however, that the House should express a view on it. It is well established that a claim in negligence for damage to property is only maintainable by a person who had either the legal ownership of or a possessory title to the property at the time when the damage occurred: Leigh and Sillavan Ltd v Aliakmon Shipping Co Ltd (The Aliakmon) [1986] AC 785 . Accepting this principle, the cargo owners argue that they meet its requirements. Rix LJ convincingly demonstrated the fallacy in the argument. He said (at 459, cols 1-2):

I would allow the appeal.

The appellants ("the shipowners") are respectively the owners and demise charterers of the bulk carrier Starsin . On 3 October 1995 they time-chartered the vessel to Continental Pacific Shipping ("CPS") for one trip, estimated 65/85 days. On 8 December 1995 the vessel began a voyage from Malaysia to Europe carrying a number of consignments of timber and plywood which had been shipped under separate bills of lading. On arrival in January 1996 it was found that 17 consignments had been damaged by fresh water; some of the cargo had been wet when loaded and negligent stowage had caused condensation which affected those and other consignments during the course of the voyage. The holders of the bills of lading ("the cargo owners") sue the shipowners for breach of the contracts of carriage contained in or evidenced by the bills of lading and in tort for negligence in stowing the goods.

Whether the shipowners are liable in contract depends upon whether they were parties to the contract of carriage made with the shippers and evidenced by the bills of lading. The time charter provided in clauses 8 and 33 that CPS should be entitled to require the master to sign bills of lading on behalf of the shipowners or authorise their agents to sign bills of lading on behalf of the master who would in turn be contracting on behalf of the shipowners. So there is no doubt that CPS had authority to cause a contract of carriage to be created between the shipowners and the shippers. But the question is whether they did so.

The port agents who signed the bills of lading used printed forms headed "Liner Bill of Lading" and bearing the name and logo of CPS. The printed text, if one reads it through carefully, shows that the forms were meant to be used by a master signing on behalf of the owners. The form has printed on the front "In witness whereof the Master of the said Vessel has signed the number of original Bills of Lading stated below". Then there is a box for the place and date of issue and, below that, another box for a signature. The form does not say expressly on whose behalf the master is signing but the master is the servant of the shipowner and unless he is authorised by the charterers to sign on their behalf and clearly does so the bill of lading will be construed as having been signed on behalf of the shipowner: see The Rewia [1991] 2 Lloyd's Rep 325, per Leggatt LJ at p. 333.

The impression that the forms are meant to be used as owner's bills is confirmed by a reading of the small print on the back. Clause 33, headed "Identity of Carrier" says in terms that the contract evidenced by the bill of lading is between the cargo owner and "the Owner of the Vessel named herein", that only the shipowner is to be liable for any breach of the contract of carriage, that the "Line, Company or Agents" which has executed the bill of lading "is not a principal in the transaction" and that "the said Line, Company or Agents shall not be under any liability arising out of the contract of carriage".

Much the same ground is covered by clause 35, which has no heading but is known as the demise clause. It says that if the vessel is not owned by the company or line by whom the bill of lading is issued "(as may be the case notwithstanding anything that appears to the contrary)" the bill of lading shall take effect only as a contract of carriage with the shipowners. It appears from an article in the Law Quarterly Review by Lord Roskill ((1990) 106 LQR 403-406) that this clause was devised to deal with conditions in the Second World War, when requisitioned ships on time charter were frequently operated by commercial liner companies which issued their own bills of lading. As the law then stood, only an owner or demise charterer could limit liability under sections 502 and 503 of the Merchant Shipping Act 1894. The words quoted above in brackets were, as Lord Roskill said, to "to put the bill of lading holder on express notice of the possibility that the ship concerned was chartered." The rest of the clause was to make it clear that in such case the owner was the carrier. Conformably with this purpose, the demise clause appears originally to have been printed on the front or "business" side of the bill of lading: see The Berkshire [1974] 1 Lloyd's Rep 185. One question in this appeal is the extent to which it remains efficacious for this purpose after its migration to the small print on the back.

The forms, therefore, were printed for use as owner's bills. But the port agents signed them, in the signature boxes on the front of the documents, "As Agent for Continental Pacific Shipping (The Carrier)" or "As Agents for the Carrier Continental Pacific Shipping" or "As Agents for Continental Pacific Shipping as Carrier". That meant, in my opinion, that anyone reading only the front of the document would think that CPS was the party assuming liability as carrier. He might have been slightly puzzled by the statement that the bill of lading had been signed by the master when it evidently had not. He may well have reflected that people often use forms which have in some respects to be adapted for the particular circumstances without deleting the inconsistent parts: see The Okehampton [1913] P 173 . But the reasonable reader of the front of the bill of lading would have had no doubt that CPS, and only CPS, was accepting liability as carrier. "Carrier" is a technical term familiar to anyone who has to deal with a bill of lading. The bill of lading evidences a contract of carriage and "carrier" is the name given to one of the parties to such a contract. As it happens, that is what condition 1(c) on the back of these bills of lading says. But that is what a reasonable reader would have thought it meant even without looking at the back. It is what Article I(a) of the Hague Rules says it means. In The Flecha [1999] 1 Lloyd's Rep. 612 Moore-Bick J., faced with a similar Continental Pacific Shipping bill of lading signed "as agents for Continental Pacific Shipping as carriers" said that the term "carrier" was being used "loosely" and that this was "not unusual or surprising". I can well imagine that a timber merchant in Kuching might say over coffee that his goods were being carried by Continental Pacific Shipping without knowing or caring whether the particular vessel was owned, demise chartered or on time charter. But such loose usage of a critical expression in the bill of lading itself does seem to me surprising.

On the other hand, a reader who turned the bill over and read the printed conditions might lose confidence in his initial impression. The identity of carrier and demise clauses would suggest deeper waters in which it might be necessary to resolve the apparent conflict between the form of signature and the other printed provisions of the bill of lading which show that it was meant for use as an owner's bill.

How is this conflict to be resolved? The interpretation of a legal document involves ascertaining what meaning it would convey to a reasonable person having all the background knowledge which is reasonably available to the person or class of persons to whom the document is addressed. A written contract is addressed to the parties; a public document like a statute is addressed to the public at large; a patent specification is addressed to persons skilled in the relevant art, and so on.

To whom is a bill of lading addressed? It evidences a contract of carriage but it is also a document of title, drafted with a view to being transferred to third parties either absolutely or by way of security for advances to finance the underlying transaction. It is common general knowledge that such advances are frequently made by letter of credit and that the bill of lading is ordinarily one of the documents which must be presented to the bank before payment can be obtained. The reasonable reader of the bill of lading will therefore know that it is addressed not only to the shipper and consignee named on the bill but to a potentially wide class of third parties including banks which have issued letters of credit.

Since a bill of lading is a legal document, the merchant or banker to whom it is addressed will know that on some questions of interpretation he will need to consult a lawyer. But he will also expect to be able to find out certain essential things for himself. These will include the identity of the carrier. The normal bill of lading recognises this distinction by having some of its terms written or printed on the front, where the businessman or banker can readily find them without a lawyer at his elbow, and the mass of other clauses printed at the back. Of course there will be cases in which the information provided on the front will be too obscure to provide the businessman or banker with the information he expects. In such a case, he may have to ask his lawyer to see whether the question can be elucidated by plunging into the small print at the back, or, if he is a banker offered the bill of lading pursuant to a letter of credit, he may simply reject it on the ground that he cannot be expected to puzzle out the answer by reference to other parts of the document. On the other hand, if the information is clearly stated on the front, the reasonable merchant or banker would go no further. The banker, for example, will accept the bill of lading when tendered against a letter of credit as having been issued by the named carrier without examining the terms on the back.

As it is common general knowledge that a bill of lading is addressed to merchants and bankers as well as lawyers, the meaning which it would be given by such persons will usually also determine the meaning it would be given by any other reasonable person, including the court. The reasonable reader would not think that the bill of lading could have been intended to mean one thing to the merchant or banker and something different to the lawyer or judge.

The proposition that bankers do not examine the contractual terms on the back of a bill of lading has long been common general knowledge and for many years the courts have said that they were not expected to do so: see Scrutton LJ in National Bank of Egypt v Hannevig's Bank (1919) 3 LDAB 213, 214 and Salmon J. in British Imex Industries Ltd v Midland Bank Ltd [1958] 1 QB 542, 551-552. In more recent times, bankers have issued public statements to this effect in the form of the ICC Uniform Customs and Practice for Documentary Credits. Article 23 of the edition which was in force when these bills of lading were issued (UCP 500) provides that if a credit calls for a bill of lading, banks will accept a document which

Article 23(a)(v) states that banks will not examine the contents of the terms and conditions of carriage, that is to say, the terms printed on the back. The position is stated even more clearly in the ICC Position Paper No. 4, published on 1 September 1994 to clarify Article 23(a)(i):

Mr Milligan QC, who appeared for the cargo owners, said that Article 23 of UCP 500 was irrelevant because it only specified the conditions upon which a bank would accept a bill of lading as conforming to the terms of a letter of credit. That had nothing to do with the interpretation of the bill of lading as between the parties (or alleged parties) to the contract of carriage. I do not agree. It is true that the purpose of Article 23, when UCP 500 has been incorporated into the terms of a letter of credit, is to specify what will count as a conforming bill of lading. But what it also shows is that, if the conditions for identifying the carrier have been satisfied, the bank will treat the document as having identified that party as the carrier. In other words, Article 23 and the Position Paper show that if a document bears upon its face the words "ABC Co Ltd as agent for XYZ Shipping, carrier [signature]" the bank will treat it as meaning that XYZ Shipping is the carrier. Since it is common general knowledge that banks almost invariable issue letters of credit on the terms of UCP 500, those terms will be part of the background available to the reasonable reader seeking to ascertain the meaning of the bill of lading. He will know that a bank, one of the potential addressees which anyone issuing a bill of lading must have in mind, would accept it as meaning that the person named on the front as the carrier was indeed the carrier. And the reasonable reader will not think that the bill of lading could have been intended to have one meaning to a bank and another to a consignee or assignee.

For this purpose it does not matter whether port agents in Malaysia are likely to have heard of UCP 500. Their knowledge and views on these matters are irrelevant because, unlike UCP 500, they are not reasonably available to everyone in the class of persons to whom the document was addressed. Nor does it matter that, as appears to have been the case here, the same port agents issued other bills in different and non-conforming form: this too is not something which the reasonable addressee of these bills could be expected to know and therefore not admissible background. The construction given to the bill of lading must be objective and uniform and, in the case of the identity of the carrier, determined by an unequivocal statement on the face of the document.

Thus it seems to me that in the present case the reasons for treating the words on the front of the bill of lading as determinative and overriding the identity of carrier and demise clauses go well beyond the general common sense rules of construction which give "preponderant importance" to the terms of the signature over the body of the document (see Universal Steam Navigation Co. Ltd v James McKelvie and Co [1923] AC 492, 500) or to written additions over boilerplate print (see Glynn v Margetson & Co [1893] AC 351 ).

I respectfully think that where the majority judgments of Sir Andrew Morritt V-C and Chadwick LJ in the Court of Appeal went wrong is that they conscientiously set about trying, as lawyers naturally would, to construe the bill of lading as a whole. In fact the reasonable reader of a bill of lading does not construe it as a whole. For some things he goes no further than what it says on the front. If the words there are reasonably sufficient to communicate the information in question, he does not trouble with the back. It is only if the information on the front is insufficient, or the questions which concern the reader relate to matters which do not ordinarily appear on the front, that he turns to the back. And then he calls in his lawyers to construe the document as a whole.

For similar reasons, I think that The Flecha [1999] 1 Lloyd's Rep. 612 was wrongly decided. Moore-Bick J, as an experienced shipping lawyer, was so conscious of the presence of the identity of carrier and demise clauses on the back of the bill of lading that could not imagine that they could be overridden by a port agent's stamp and signature on the front. He said that the term "carrier" on the front was used loosely but I think that what he really meant was that it must have been a mistake. If one were construing the document as a whole, giving equal weight to all its clauses, one might possibly conclude that it was a mistake: compare Mannai Investment Co. Ltd v Eagle Star Life Assurance Co. Ltd [1997] AC 749 . But the bill of lading is addressed, among others, to persons who will try if possible to identify the carrier simply by what it says on the front. What The Flecha and the majority decision in the Court of Appeal did was to drive a wedge between the reasonable perception of a bank taking up the bill and the construction which would later be given to the bill by a court in litigation, perhaps involving the same bank. This does not seem to me commercially fair.

Mr Milligan argued that in the alternative that the description of CPS as carrier in the signature box was not inconsistent with the shipowners being parties to the contract of carriage. In Fred. Drughorn Limited v Rederiaktiebologet Transatlantic [1919] AC 203 the House of Lords held that a charterparty which described one of the parties as "charterer" was not inconsistent with his being agent for an undisclosed principal who also assumed the rights and duties of charterer. So, in this case, Mr Milligan said that the description of CPS as "carrier" was not inconsistent with giving effect to the demise clause under which CPS were deemed to be agents for the shipowners. They were disclosed agents (the agency being disclosed by the demise clause) for an unnamed principal.

This might be a powerful argument if the bill of lading were a different kind of document. But, for the reasons which I have given, I think that if the carrier is plainly identified by the language on the front of the document, one never gets to the demise clause on the back. The language on the front simply takes priority and no attempt at reconciliation is required. Mr Milligan also submitted that CPS may have contracted both for themselves and the shipowners, the latter being unnamed or undisclosed principals. Rix LJ, who appears to have floated this theory in the Court of Appeal, said that it might be considered "novel and inconsistent with the settled expectation of the shipping trade". He would know this better than I, but I do not think that any reasonable merchant or banker who might be assumed to be the notional reader of this bill of lading would imagine that there was more than one carrier or that the carrier was anyone other than CPS.

It follows that in my judgment the appellants were not parties to the contract of carriage and are therefore not liable in contract.

Are they liable in tort? There is no dispute that the damage to the cargo was caused by the negligence of their servants employed on the vessel. The case has been argued on the basis that the shipowners owed the normal duty of care to the cargo owners on the ground that it was foreseeable that bad stowage would damage their goods. The existence of that duty could be reinforced by the fact that the shipowners were bailees or sub-bailees of the goods.

But a person who sues for damage to goods must show that he had title to the goods at the time the damage occurred. Otherwise he has suffered economic damage rather than physical damage to his property: he has paid for goods which were damaged and therefore worth less at the time that he acquired title. Such loss cannot be recovered in an action for negligence: The Aliakmon [1986] AC 785 .

In the present case, only Makros Hout had obtained title before the voyage began. Prima facie, therefore, they are entitled to sue. The other cargo owners obtained title from the shippers at various stages during the voyage. The judge held that notwithstanding that the negligent act was the stowage at the commencement of the voyage and that, once the voyage had begun, progressive damage to the cargo was inevitable, the other cargo owners could recover for the proportion of the damage which had been caused to their consignments after they had obtained title.

My Lords, in agreement with the Court of Appeal, I think that this was to treat the progress of the damage as creating new causes of action which accrued per diem in diem . But in my opinion there was a single cause of action which accrued to the persons who owned the cargo at the time when the negligent stowage caused it any significant damage. That cause of action comprised all damage caused by the negligent stowage, even if some of that damage did not manifest itself until after they had parted with ownership. As significant damage was suffered by all the cargo before title passed to the cargo owners (other than Makros Hout), they have no cause of action in tort.

The respondents relied upon the decision of the House of Lords in Darley Main Colliery Co v Mitchell (1886) 11 App Cas 127. But that was an unusual case in which the cause of the damage (digging coal under ground) was not a wrongful act. It gave rise to a cause of action only in so far as it let down some part of the surface. So there was no unifying element in the cause of action such as, in this case, is provided by the negligent stowage. Each letting down of the surface was a separate cause of action. In the present case, all damage caused by the negligent stowage is a single cause of action which is complete once any significant damage has occurred.

That means that only Makrous Hout have a prima facie claim in negligence. But the shipowners say that even that claim is excluded by an exemption clause on the back of the bill of lading, namely condition 5, known as the Himalaya clause after the decision in Adler v Dickson (The Himalaya) [1955] 1 QB 158 . As sub-bailees, they may also have been entitled to rely on any terms of bailment contained in the contract between them and the charterers as bailees, i.e. the charterparty: see The Pioneer Container [1994] 2 AC 324 . But no reliance has been placed by either side on those terms. The argument has turned entirely upon the effect of the Himalaya clause. The relevant provisions are quoted in the speech of my noble and learned friend Lord Bingham of Cornhill and I shall adopt his numbering of the relevant parts.

A Himalaya clause in a contract of carriage is designed to create contractual relations between the shipper and any third parties whom the carrier may employ to discharge his obligations. It does so without infringing the English doctrines of privity of contract and consideration, which, until the Contracts (Rights of Third Parties) Act 1999, prevented third parties from claiming benefits under contracts. The way it works is this. The shipper makes an agreement through the agency of the carrier with the third party servant or contractor. Such third parties may have authorised the carrier in advance to contract on their behalf or they may afterwards ratify the agreement. The terms of the agreement are that if such a third party renders any services for the benefit of the cargo owner in the course of his employment by the carrier, he will be entitled to the exemptions and immunities set out in the clause. At that stage, the agreement is not a contract. The third party makes no promise to the shipper to render any services and, until he has actually rendered them, no contract has come into effect. It is the act of rendering the services which provides the consideration and brings into existence a binding contract under which the third party is entitled to the exemptions and immunities. The efficaciousness of the clause to achieve these results has been affirmed by the decision of the Privy Council in The Eurymedon [1975] AC 154 . The theory of the agreement which becomes enforceable conditionally upon the act providing consideration was developed by Sir Garfield Barwick CJ in his dissenting judgment in the High Court of Australia in Port Jackson Stevedoring Pty Ltd v Salmond and Spraggon (Australia) Pty Ltd (The New York Star) (1978) 139 CLR 231 and adopted by the Privy Council when it affirmed his judgment on appeal: see The New York Star [1981] 1 WLR 138 .

The first question is whether this version of the clause created the necessary agency. I agree with my noble and learned friend Lord Bingham of Cornhill that words have been omitted because of a common copyist's error and that it is possible to identify the substance of the missing words. The clause ought therefore to be read as if it contained them. They make it clear that the carrier acts as agent to contract for the exemptions in parts (1) and (2). I do not think that there is any contradiction between the somewhat sophisticated technique of construction which has to be applied to restore clause 5 to its intended form and the more rough and ready approach used to decide who is the carrier. The latter is, as I have said, a question which the reasonable merchant or banker would expect to be able to decide for himself. On the construction of condition 5, he would undoubtedly want to consult a lawyer. For similar reasons, I do not think that it is relevant that the Himalaya clause is preceded by a separate and verbose sentence purporting to confer upon the carrier a series of immunities of which some are probably in conflict with the Hague Rules and others unnecessary.

The next question is the nature of the exemption conferred by part (1). It confers immunities upon the carrier's servants, agents or independent contractors. I agree with Colman J. and the unanimous Court of Appeal, for the reasons they gave, that the shipowners were independent contractors to carry the goods for the charterers. The cargo owners no longer contest this point. The condition then says that no such servants, agents and independent contractors "shall in any circumstances whatsoever be under any liability whatsoever to the shipper". That seems reasonably clear. It appears to exempt them from any liability. But Colman J. and all the members of the Court of Appeal agreed that it did not. Why not?

First it was said that part (1) is not a contract between the shipper and a third party. It is a covenant with the carrier, enforceable only by the carrier, by which the shipper promises that neither he nor his successors in title will sue the third party. The way to enforce it is for the carrier to obtain an injunction prohibiting the shipper from suing the third party: see The Elbe Maru [1978] 1 Lloyd's Rep. 206. In the Court of Appeal, Chadwick LJ ( at p. 471) made his position on this point very clear: "The starting point", he said, "is to appreciate that [part (1)] has effect only as an agreement between the shipper and the carrier." He referred to the words of part (3) ("all such persons shall to this extent be deemed to be parties to the contract contained in or evidenced by this Bill of Lading") and said that the words which he italicised showed that the agency mechanism was to apply only to part (2).

I have two difficulties about this argument. The first is that, read as a whole, the Himalaya clause makes it clear that the carrier contracts as agent for the third party in respect of both parts (1) and (2). The agency mechanism is not in my opinion created by part (3); it is created by the words in part (2) -

Part (3) has a different function which was in my opinion correctly identified by Beattie J. in his judgment at first instance in The Eurymedon [1971] 2 Lloyd's Rep. 399, 409:

The purpose of deeming the third party to be party to the bill of lading to that extent is to ensure that the exemption clause is enforceable against subsequent assignees. As part of the bill of lading it will be enforceable under the Carriage of Goods by Sea Act 1992 or possibly the common law rule that the consignee is bound by his acceptance of the bill of lading and request for delivery of the goods: see The Eurymedon [1975] AC 154 , 168.

It may be that Colman J. and the Court of Appeal were misled because they were not given the words which had dropped out of the clause. But even without them I would have decided that the agency covered both parts (1) and (2). That is because in my opinion, and I understand it to be that of all your Lordships, part (1) simply cannot be read as a covenant not to sue. It does not say that the shipper is not to sue the third party. It says that he shall not be under any liability. A similar point arose in Gore v Van Der Lann [1967] 2 QB 31, in which the plaintiff contracted with the Liverpool Corporation that their bus drivers would not be liable to her for any damage however caused. The Corporation did not purport to contract as agent for the drivers and when the plaintiff sued a driver it was conceded that he could not rely upon the exemption: see Midland Silicones Ltd v Scruttons Ltd [1962] AC 446 . But the Corporation intervened to claim that the contract was a covenant not to sue the drivers which they could enforce. The Court of Appeal rejected this argument, saying that the agreement could not be construed as containing such a covenant. So it seems to me that part (1) is either a contract of exemption between shipper and third party or it is wholly ineffective. In my opinion it was the former.

The second argument is that if the first part of the clause means what it appears to say, cases like the The Eurymedon [1975] AC 154 , The New York Star [1981] 1 WLR 138 and The Mahkutai [1996] AC 650 would have been decided on different grounds or at any rate argued differently by the eminent counsel who appeared in them. This kind of argument is always rather speculative - one seldom knows exactly why some point was not taken - and a close examination of those three cases does not make it any stronger. When The Eurymedon was decided at first instance in New Zealand ([1971] 2 Lloyd's Rep. 399) Beattie J. had no doubt (at p. 408) that the first part of the clause meant what it said:

Beattie J. noted that the first pleaded defence was simply that the defendants (in that case, stevedores) were not liable because the work was done in the course of their employment. The second defence pleaded a package limitation and the third the time bar under the Hague Rules. The judge noted that all three defences depended upon showing that the agency mechanism succeeded in creating contractual relations between the defendants and the shipper and that if that worked, the time bar would be enough to defeat the action. But he observed in passing that -

In other words, the only answers which Beattie J. could see to a complete exemption under the Himalaya clause were (1) the agency mechanism did not work or (2) the contract created by the agency mechanism was a contract of carriage, so that the exemption was struck down by Article III.8 of the Hague Rules. Since the decisions of the Privy Council in The Eurymedon and The New York Star it is clear that in principle the agency mechanism can work and I have already explained why I think it works for part (1) of this clause. That leaves the question of whether the purported exemption would be struck down by the Hague Rules. I shall come back to that point later.

In the New Zealand Court of Appeal, Turner P. summarised the issue as follows:

Perry J. clearly held the same view, because he remarked:

The New Zealand Court of Appeal allowed the appeal on the ground that they did not think that the agency mechanism worked. They could not find any consideration moving from the stevedores. So they did not consider the point which Beattie J. said would arise if the agency did work but there had been no time bar, namely whether the contract was subject to the Hague Rules.

In the Privy Council it was recognised, as it had been by Beattie J., that if the agency mechanism worked, the time bar would be enough to enable the appellants to succeed. So nothing was said about the potential effect of the exempting part of the clause.

In The New York Star [1981] 1 WLR 138 there was a general exemption clause and a separate time bar. The stevedores recognised that there was a strong argument that the doctrine of fundamental breach might prevent the exemption clause from applying to the facts. The stevedores had handed over the goods to a thief without requiring production of the bill of lading. (Compare Sze Hai Tong Bank Ltd v Rambler Cycle Co. Ltd [1959] AC 576 .) In the Privy Council Mr Gleeson QC, for the stevedores, included reliance upon the "general exclusion provisions" as an alternative at the end of his printed case (paragraph 20) but his stronger argument, upon which he succeeded, was based on the time bar. So there was no discussion of the issues which arise in this case.

The Himalaya clause in The Mahkutai [1996] AC 650 was said by Rix LJ (at p. 461) to be similar but not identical. It was similar in that it had words similar to part (2), although in a somewhat narrower form. But it was different in that it did not have part (1) at all. Instead, it had what was plainly a covenant not to sue in a form similar to that in the The Elbe Maru [1978] 1 Lloyd's Rep. 206. There was no general exemption clause. This difference seems to me rather important and makes it less surprising that the case contains no discussion of the effect of a general exemption clause. It is true that Lord Goff of Chieveley said ( [1996] AC 650 , 666) that the function of a Himalaya clause was -

But this dictum is clearly addressed to the clause used in the case. I do not think it would be fair to treat it as laying down that this was the only function of any Himalaya clause.

The third argument is from redundancy. The second part of the clause says that servants and independent contractors are to have the same protection as the carrier. So, it is said, if the first part exempted the independent contractor from all liability whatever, he would not need the protection of the second part. This argument was accepted by Colman J. ([2000] 1 Lloyd's Rep. at p. 99) and by Rix LJ and Sir Andrew Morritt V-C in the Court of Appeal ([2001] 1 Lloyd's Rep. at pp. 462, 476).

My Lords, I seldom find arguments from redundancy very compelling and I think that in the case of a Himalaya clause they carry little weight. I do not think it surprising that when the clause was drafted (probably after Adler v Dickson (The Himalaya) [1955] 1 QB 158 ) the draftsman thought it might be prudent to wear belt as well as braces. In the legal climate of that time such prudence would have been justified. Exemption clauses were under threat from extensions of the fundamental breach doctrine and the draftsman might have thought that if the exemption in part (1) failed, the third party should at least be able to fall back upon the exemptions and time limits available to the carrier. That is what happened in The New York Star [1981] 1 WLR 138 .

That brings me to the fourth argument, which is that the complete exemption conferred by part (1) is cut down by Article III.8 of the Hague Rules, which provides that any clause in a contract of carriage relieving "the carrier or the ship" from liability for negligence shall be null and void. I confess that on this point my opinions have fluctuated but in the end I have been persuaded that the reasoning of Lord Hobhouse of Woodborough is correct and that Article III.8 does have this effect.

Putting the argument in my own words, it seems to me to run as follows. I do not think that the collateral contract between shipper and independent contractor is a "contract of carriage" so as to attract the application of the Hague Rules. But part (3) says that the independent contractor "shall to this extent be deemed to be parties to the contract contained in or evidenced by this Bill of Lading". That means, as I said earlier, that he is a party only for the purpose of taking the benefit of the exemption clause against the shipper and any transferee of the bill of lading. But, for that purpose only, the provisions of the bill of lading, insofar as they are relevant, apply to him. The only provision which has been suggested as relevant in the present case is Article III.8, which applies by virtue of the paramountcy provision in part (2). That does apply to exemption clauses and restricts their effect. I think that this argument was also accepted by Colman J. ([2000] 1 Lloyd's Rep. 85, 100) and Rix L.J. ( [2001] 1 Lloyd's Rep 437 , 462).

I think that such a construction is in accordance with the general policy of the Hague Rules, which was to provide an acceptable balance in distributing the risks of loss and damage between carrier and cargo owners. The rules are intended to preserve the common law remedies which cargo owners would have in English law for loss of or damage to the cargo in the circumstances there specified. Of course they do have this effect anyway in relation to the contractual carrier, the charterer. But the preservation of these remedies must also be considered in relation to the procedures available for enforcement. By section 20(2)(g) of the Supreme Court Act 1981 the High Court has admiralty jurisdiction in respect of claims for loss of or damage to goods carried in a ship (which may, as in this case, be a claim in tort) and by section 20(2)(h) it has jurisdiction in respect of any claim arising out of any agreement relating to the carriage of goods in a ship. Jurisdiction may be founded by the arrest of the ship, which will also provide security for the claim. Without the right to these enforcement procedures, a cargo claim against a foreign carrier would in many cases be unenforceable. But by section 21(4)(i), the action in rem may be brought against the ship only if the person who would have been liable in personam was at the time when the action was brought the beneficial owner or charterer by demise.

It follows that in the case of goods carried under a bill of lading issued by a time charterer, the liability in personam of the time charterer will not enable the cargo owner to arrest the ship. He may do so only if the shipowner is also liable in tort. But such liability will not exist if the time charterer is able to stipulate for complete exemption on the part of the owner. The remedy which the Hague Rules were intended to preserve may in such cases be unenforceable.

I would therefore make the orders proposed by my noble and learned friend Lord Bingham of Cornhill.

The Appellants are companies respectively registered in Liberia and Singapore. At all material times, the former was the owner and the latter the demise charterer of the motor vessel Starsin. No distinction has been drawn between them and I will refer to them simply as the shipowners. The Starsin was a bulk carrier of about 27,000 tons deadweight capacity. She flew the flag of St Vincent and the Grenadines. She had 5 holds and 9 hatches and was crewed by Ukrainian officers and men.

On 3 October 1995, the shipowners time-chartered the Starsin to a company called Continental Pacific Shipping Ltd of Douglas in the Isle of Man. The timecharter was expressly governed by English law. The NYPE form was used. The duration of the charter was to be one timecharter trip (estimate 65/85 days). Under this form of timecharter, the ship remains in the possession of the shipowners and the master and crew remain the shipowners' servants (cl.26). The timecharter contained a number of clauses dealing expressly with the issuing of bills of lading: "the Captain ... is to sign bills of lading for cargo as presented in conformity with Mate's or Tally Clerk's receipts" (cl.8); "the Master to authorise, if required, the Charterers or their Agents to sign bills of lading on his behalf, provided in conformity with Mate's Receipts" (cl.33). As regards the terms of the bills of lading, the timecharter stipulated that certain clauses were both to be deemed to be incorporated in the timecharter and to be included in all bills of lading issued under the timecharter. Clause 56 provided:

This was the charter party under which the vessel was operating when she called at various Malaysian ports in the following year and loaded various consignments of timber and timber products. It appears that at some ports some timber was loaded which had been wetted by rain before shipment. It was loaded nevertheless, notwithstanding the fact that the presence in the holds of wet cargo loaded at earlier ports might make the vessel unfit ( ie , unseaworthy) for the carriage of timber to be loaded at a later port (Hague Rules, Art.3 r.1) and/or the loading of wet timber might constitute a breach of duty properly to care for consignments loaded earlier (Art.3 r.2). Clean shipped on board bills of lading were however issued. But the cargo was in any case improperly stowed, without sufficient dunnage to separate the consignments, enable the cargo and the holds to be adequately ventilated and to protect the cargo from being wetted by ship and cargo sweat ( ie , condensation). After loading at the various ports, the vessel set out on her voyage to Northern Europe on 8 December 1996. She arrived off Antwerp on about 19 January 1997 and discharged her cargo at ports which included Rotterdam and Avonmouth. Some 17 consignments out-turned seriously damaged by fresh water. It is these consignments which form the subject matter of these actions. The actions were commenced in the Admiralty jurisdiction of the High Court, two in rem , and a third in personam . The shipowners entered appearances in all actions. The charterers although they were sued as well did not appear and have taken no part in the actions. Your Lordships were told that they were now insolvent but it does not follow that this has any relevance having regard to the terms of the timecharter.

The claimants were various merchants who had acquired the title to the relevant consignments and to whom the relevant bills of lading had been endorsed. They were the 'notify parties' at the relevant discharge ports but not the consignees named in the bills of lading. The consignees were either 'order' or the sellers' bankers through whom the documents were to be negotiated. All the bills of lading issued for the voyage to Europe were on the forms used by 'Continental Pacific Shipping', a trading name which does not distinguish between the charterers and their sister company Continental Pacific Shipping NV of Antwerp. This ambiguity cannot be resolved by reference to the bills of lading themselves, only by having regard to other documents, eg the timecharter. Assuming that the usual practice was followed, the shippers' agents would have purchased these forms from stationers or the port agents at the various loading ports. They do not all have exactly the same wording; they contain a considerable number of mis-prints and omissions, not the same ones at every port. None of these defects and variations are material until I get to the 'Himalaya' clause issue and I will ignore them. Thus, the bill of lading forms included, besides the "Basis of Contract" clause (clause 2) contractually incorporating the Hague Rules, to which I will come later, the following:

On the front of the form, the heading is "Liner Bill of Lading No. "; there is a flag or logo and the words "Continental Pacific Shipping". At the foot there is a box saying "Signature". The front and back of the form are therefore unequivocal. The front is to be signed by or on behalf of the Master, that is to say, on behalf of the employer of the master, the shipowner (or possibly a demise charterer of the vessel). The definition in clause 1(c) confirms this and in no way detracts from it. Finally, clause 33 expressly provides that the contract evidenced by the bill of lading is to be with the owner of the vessel to the exclusion of anyone else. Clause 35 stipulates for the same outcome. All this is incontrovertible.

The complications in the present case arise from the fact that, whereas, for a large majority of the consignments shipped on the Starsin, the bills of lading were signed by the port agents simply "as agents", the bills of lading for the consignments the subject of these actions were signed by the port agents as agents "for The Carrier Continental Pacific Shipping" or "for Continental Pacific Shipping (The Carrier)" or "Continental Pacific Shipping As Carrier". For the first of these versions, it appears that all the words are typed; in the second and third, it appears that a rubber stamp was used. Hence the first question raised on the appeal. Are the shipowners liable in contract on these bills of lading, so signed? Colman J and Rix LJ held that they were not but the majority in the Court of Appeal, Morritt V-C and Chadwick LJ, held that they were. No point was taken on the claimants' title to sue on the bills of lading under the Carriage of Goods by Sea Act 1992 (the successor of the Bills of Lading Act 1855). No other contractual claim was made by the claimants against the shipowners at the trial. But, if the claimants could establish this contractual cause of action, they would, by virtue of the 1992 Act, not need to prove what damage had been caused at what stage of the ocean voyage.

The rival arguments were essentially very short. The claimants, through Mr Milligan QC, argued that there was no necessary contradiction between words of description on the front and the unequivocal provisions of clause 33 and that clause 33 should be applied, as should clause 35. They further submitted that if there was any inconsistency it could be resolved by reading the signature as still being for the master and merely lengthening the chain of agency; the local agents signed as agents for the timecharterers who were to be taken as being the agents for the shipowners. He used the analogy of 'undisclosed' principals. The difficulty for Mr Milligan was that he did not seek to prove any such undisclosed agency as a fact and had to make good his argument, if he could, as a matter of the construction of the document. He also pointed out that there were other clauses besides clauses 33 and 35 in the standard terms which would require at the least amendment and adjustment on the shipowners' argument.

Mr Gee QC for the shipowners argued that the typed or stamped words added to the box containing the signature and the designation of 'Continental Pacific Shipping' as the carrier tied in with the definition of 'carrier' in clause 1(c); he was also prepared, at this point in his argument, to pray in aid the provision of Article I(a) of the Hague Rules that "'carrier' includes the owner or the charterer who enters into a contract of carriage with a shipper". He argued that the signature was unequivocal. It showed that the specific intention of the local agents was that Continental Pacific Shipping were the carrier and that this implicitly over-rode the other parts of the form, front and back, including clause 33 and there was no need to do more.

Mr Gee also sought to develop his argument in a number of ways. He referred to the Uniform Custom and Practice for Documentary Credits and the ICC position papers which amplify them. The Uniform Custom is essentially a bankers' code of practice prepared for the purpose of clarifying how documentary letters of credit should be drafted, how the purchaser's stipulations should be given effect to and what are the respective rights and liabilities of the opening bank, the confirming bank and the notifying bank. It is as a matter of the standard practice of banks contractually incorporated into documentary letters of credit. He rightly pointed out that the current version of the code (UCP500) has made significant changes to the previous versions and among other things, requires that the actual identity of the carrier should be disclosed and particularised by naming it or him on the face of the bill of lading. Thus, if it is a master's bill of lading, the actual name of the master must appear on the face of the bill of lading; or if it is the shipowner or a charterer a specific identification of the relevant legal entity must be made. All this is a worthy aspiration and it remains to be seen how deep the new banking practices will penetrate into the former habits of port agents in places such as the East Indies. In the present case they did not. None of the bills of lading issued in respect of any of the cargo carried from Malaysia to Europe were fully compliant. The bills of lading the subject of these actions failed to identify which Continental Pacific entity was being referred to. There was no evidence to suggest that the different signatures to the bills of lading the subject of this action derived from any attempt to comply with the code. The value of drawing attention to the code is that one should now be prepared to look for the identity of the carrier on the face of the bill rather than in the clauses on the reverse. Bills of lading are transferable documents and can be expected, as often as not, to form part of the documentation to be tendered under a documentary letter of credit.

Further arguments of Mr Gee had less relevance. He sought to discredit clause 35 as being a version of the 'demise clause' which was historically obsolete. However it is a standard clause to include in very many types of bill of lading and one would have been surprised not to find it in these bills of lading. But, besides, these bills of lading also include an "IDENTITY OF CARRIER" clause, clause 33, expressly addressing and answering the relevant question of which no criticism is made. He also sought to decry the smallness of the print. This is an old argument which has never been accepted in this context: Elderslie v Borthwick [1905] AC 93 ; Scrutton on Charterparties and Bills of Lading, Article 9. It can be commented that his criticism did not inhibit him from relying at this and later stages of his argument upon clauses in equally small print which assisted his case.

In my judgment the salient fact is that the signatures contradict the form. The signature is not neutral or equivocal, nor, for that matter, is the form. Where the (original) parties, by issuing and accepting these bills of lading with these signatures, have expressly agreed that Container Pacific Shipping ( sic ) shall be the contracting party, they have implicitly agreed that inconsistent clauses will be overridden. The special words, typed or stamped, placed in the signature box demonstrate a special agreement. Effect must be given to that agreement. The contracts contained in these bills of lading are contracts with Container Pacific Shipping, now identified by reference to the timecharter as being the limited liability company in the Isle of Man.

There are two observations to made about this conclusion. The first is that 'shipped on board' bills of lading, as all these bills of lading were, will normally have been preceded by some anterior contract ordinarily on the terms of the bill of lading form to be used. By definition the bills of lading will not then have been signed and issued. If the consignment is dropped in the course of loading, as happened to the fire tender in Pyrene v Scindia [1954] 2 QB 402, a bill of lading may never be issued but the contract will still be one which is "covered by a bill of lading" under Article I(b) of the Hague Rules ( ib ). The operation and the claim of the goods owner is subject to the contemplated bill of lading and the Hague Rules. Transposing the situation to that existing in the present case, to accept the shipowners' argument raises the possibility that the signatures changed the pre-existing contracts. However, there is no evidence or finding on this aspect so it cannot assist either side on the appeal. The second observation is that the claimants are subsequent holders of the bills of lading by endorsement. Their contractual rights must be ascertained by reference to the bill of lading document itself. Look at the bill of lading, front and back, and arrive at a conclusion. That is what I have done and my conclusion is that the contract is not with the shipowners.

The appellant shipowners therefore succeed on the first question. but that is not the end of the case. The parties have agreed that three further issues arise:

The parties' formulation of issue (2) substantially represents the present position of the parties and in this form is no longer in issue between them. But before the judge, the shipowners had argued that there had been a bailment on terms. The judge decided the case upon the tort of negligence and the ordinary duty of care of any person not to cause foreseeable physical damage to another's property (or person) by unreasonable conduct: ie the Donoghue v Stevenson duty of care. For obvious tactical reasons Mr Gee did not seek before your Lordships to establish a bailment on terms, or any bailment; he contented himself with the bald assertion that any such analysis was simply wrong. Nevertheless, it is necessary, in order to explain how issue (3) comes to be in the form which it is and to discuss the important fourth issue, to consider the role of bailment and sub-bailment.

A contract of carriage (unlike a contract for carriage or a contract of affreightment) is a contract of bailment. A bill of lading fulfils a number of roles. It is a receipt; it is evidence of a contract of carriage; it is a transferable document of title. Its primary role is as a receipt for the goods at the port of shipment. Thus these bills of lading have on their face the name of the vessel and the particulars of the goods as furnished by the merchant followed by the words

So far I have only been talking about the position of the timecharterers. They are the persons on whose behalf these bills of lading have been issued. But these bills of lading by stating (as was the case) that the consignments had been shipped on board the Starsin as the carrying vessel, confirm that there has been bailment of the goods to the shipowners, either direct from the shippers to the shipowners as in the Elder Dempster case, or, according to the view I prefer to take of the present case, by way of a sub-bailment by the timecharterers to the shipowners . So, on my preferred view, there is a bailment by the shipper to the timecharterers and a sub-bailment by the timecharterers to the shipowners. The shipowners having the possession of the goods as sub-bailees, what liabilities on their part does this give rise to vis-à-vis the owners of the goods and what are the terms of their sub-bailment? Both these questions were authoritatively answered in the unanimous judgment of the Privy Council delivered by Lord Goff of Cheiveley in The Pioneer Container [1994] 2 AC 324 .

The case arose out of a collision off Taiwan. One of the ships was sunk. It was carrying cargo which it had loaded in Taiwan and was carrying to Hong Kong. The owners of some consignments of cargo commenced an admiralty action in Hong Kong and arrested a sister ship of the vessel which had sunk. The owners of that ship sought a stay of the Hong Kong proceedings relying upon a Taiwan exclusive jurisdiction clause contained in their bills of lading. But two groups of cargo owners had no contractual relationship with the shipowners. The bills of lading issued to them had been issued by others and covered carriage from, in one case, the USA to Hong Kong and, in the other, from Taiwan to Europe. Each of those bills of lading contained a clause: "The carrier shall be entitled to subcontract on any terms the whole or any part of the ........ carriage of the goods and any and all duties whatsoever undertaken by the carrier in relation to the goods." These bills of lading did not contain any exclusive jurisdiction clause but did contain a 'Himalaya' clause (see the arguments at pp.327-332 and Lord Goff at p.344). The actual carriers from Taiwan to Hong Kong were subcontractors of the persons who had issued the bills of lading in the hands of the cargo owners and it was to those persons that the actual carriers had issued their bills. How then could the actual carriers have a right to enforce the exclusive jurisdiction clause against the cargo owners?

Lord Goff, having referred to the obvious advantages of concentrating all litigation arising out of such an accident in a single jurisdiction and to the doctrine of privity of contract as reaffirmed by your Lordships' House in Midland Silicones v Scruttons [1962] AC 446 , said (pp.335-342, 344):

To summarise, this judgment establishes that:

The upshot is that the claimant cargo owners (except for Hunter Timber Ltd, who could not prove when they had become the owners of 'their' consignments) would have been entitled to hold the shipowners liable as bailees or sub-bailees and the shipowners would have been entitled to rely upon the terms upon which they had taken the goods into their possession. For that purpose, it is the contract between the bailees and the sub-bailees which must be looked at, not that between the cargo owners and the bailee (unless a question has arisen as to the bailee's authority to sub-bail, which is not the case here). The relevant contract is the timecharter. The timecharter stipulates, for the protection of the shipowners, that the Hague Rules are to be incorporated into the contract between the shipowners and the timecharterers. It also stipulates that a Clause Paramount shall be included in all bills of lading.

Further, by way of comment, the cargo owners could, by relying upon the goods owner/sub-bailee relationship, have put the burden on the shipowners to excuse their failure to deliver the goods undamaged. But at the trial they did not. They relied upon a Donoghue v Stevenson claim and had to discharge the burden of proof which that entailed.

The claimants did not adduce at the trial evidence to prove the timing and cause of damage to the individual consignments or indeed the separate consignments of each claimant. They accordingly did not obtain more than global findings from the judge as to what, on the balance of probabilities, was the progress of the damage over all the claimants' consignments as a whole and without distinguishing between them. Before the judge only the Hunter claimants failed to recover. They were unable to prove that they had acquired the title to the consignments in respect of which they were claiming at any time before the end of the voyage. They were in no better position than the claimants in The Aliakmon [1986] AC 785 . In the Court of Appeal, Rix LJ pointed out that the same reasoning applied as well to defeat the Donoghue v Stevenson claims of the other claimants, except Makros Hout BV. They could not claim for bare economic loss. They had to prove what actual damage had been done to their consignments after they had acquired their title to them. Having failed to do so they could not recover anything. Rix LJ was clearly correct. The cross-appeal of the two unsuccessful claimants in the Court of Appeal, Fetim BV and Homburg Houtimport BV, should be dismissed. That leaves Makros Hout. They were able to prove that they had acquired their title to their consignments before 8 December 1995 which the judge had found was the date upon which consignments started to suffer significant damage. Macros Hout were therefore entitled to recover in tort subject to the shipowners being able to set up a contractual defence.

The shipowners submit that, as the actual carrier, they are entitled to be exempt from any liability whatsoever in respect of this cargo, ie they assert a total exemption from any legal liability no matter what breaches of the sub-bailment they may have committed. They are not content that they should have the benefit of the same exceptions and limitations as are available to the 'contracting' carrier (for the obvious reason that those exceptions would not enable them to defeat the claim). They deny any contractual or other application whatsoever of the Hague Rules to them. These are remarkable submissions which seek to carry the reach of a 'Himalaya' clause in a bill of lading far further than any previous decision. If their submissions are correct they are highly significant. They will provide the actual performing carrier with a route for evading by means of a bill of lading clause the Hague Rules scheme. It will put English law at odds with those legal systems which are not based upon privity of contract and where the Hague Rules as amended by later international instruments take direct legal effect as part of their commercial or maritime codes ( ie do not have to go through a contractual gateway): see for example Article 437 of the German Commercial Code quoted below. They run counter to the Hague-Visby Rules, Article IV bis, made part of British law by the Carriage of Goods by Sea Act 1971 and Article 10 of the United Nations Convention on the Carriage of Goods by Sea 1978 (the Hamburg Rules).

The Hague-Visby Rules were the result of a protocol up-dating the Hague Rules Convention prepared by the Brussels Diplomatic Convention on Maritime Law and signed there in 1968, since when there have been extensive ratifications and accessions. It was enacted into English law in place of the 1924 Act by the 1971 Act. For present purposes, it had two important features. First it made its application a matter of law not just a matter of contract, thus rendering the peculiarly Anglo-Saxon contractual problems largely irrelevant. S.1(2) of the 1971 Act provides: "The provisions of the Rules, as set out in the Schedule to this Act shall have the force of law." (See also The Hollandia [1983] 1 AC 565 at 572, per Lord Diplock.) Secondly, in a new Article 4 bis it expressly addressed the servant and agent liability problem. Article 4 bis having in r.1 confirmed that the defences and limits of liability provided for in the Rules should apply in any action against the carrier in respect of loss of or damage to goods covered by a contract of carriage whether the action be founded in contract or in tort, went on to provide in r.2

The Hamburg Rules were adopted by a United Nations conference on the carriage of goods by sea at Hamburg in 1978. They made many changes to the previous Rules. Some of them were not well drafted and did not command the same wide support and degree of adoption as the Hague-Visby Rules. However the provisions relating to the actual carrier were hailed by at least one well informed commentator, William Tetley QC, as a "major advance in the law" (1979 LMCLQ 1) and have for example been incorporated in the German Commercial Code. Framework legislation for their enactment was passed in Australia (1991) and Canada (1993) but in neither country has it been taken any further nor has the United Kingdom taken any such step. For present purposes the relevance of the Hamburg Rules is that the international shipping community was again not prepared to support the immunity for which the shipowners contend in the present case. The Hamburg Rules introduce a specific definition of "Actual Carrier" as being (in brief) "any person to whom the performance of the carriage of the goods has been entrusted .... by the carrier ..." (Art 1 (2)) There is an equivalent of the Hague-Visby Article 4 bis. (Art 7) There is a new Article 10 which deals specifically with the liability of the actual carrier and provides that "all the provisions of this Convention governing the responsibility of the carrier also apply to the actual carrier for the carriage performed by him". (Art 10 (2)) Rules 3 to 6 of this Article then go on to identify that the liability of the actual carrier and the (contracting) carrier is joint and several (4), but no more than the Rules monetary limits shall be recovered in the aggregate (5), rights of recourse between the (contracting) and the actual carrier shall not be affected (6) and any surrender or waiver of rights by the (contracting) carrier shall not affect the rights of the actual carrier unless agreed to by him in writing (3). Thus, Article 437 of the German Commercial Code provides, in translation:

As regards other jurisdictions, the internationally accepted version of the Hague-Visby Rules have been adopted by the majority of major trading nations either specifically as have the United Kingdom, Australia, Canada, Hong Kong and New Zealand, or indirectly by simply incorporating their terms into their commercial or maritime codes. This was, for example, the position in relation to the Hague Rules in Belgium, Holland, Germany, France and many other European countries. According to Tetley ( loc cit ), France and some other countries have also introduced express provisions in their commercial codes to define the liabilities of stevedores. The contractual subtleties do not arise; the shipowners' liability is determined by the provisions of the relevant commercial code. It is also noteworthy that the Eurymedon type 'Himalaya' clause is not the only type of bill of lading clause that came into use to address the question of servant/agent/independent contractor liability. In the United States of America, a clause came into use which simply used an expanded definition of "carrier". (See The Mormacstar [1973] 2 Lloyds 485, US 2nd Circuit, in which the effectiveness of such a clause was upheld.) This route did not provide greater exceptions than those available to the 'contracting' carrier.

Before examining the decided cases and the principles which they disclose, it is as well to bear in mind a basic rule of construction of contracts of carriage. If a party, otherwise liable, is to exclude or limit his liability or to rely on an exemption, he must do so in clear words. Unclear words do not suffice. ( eg. The Pera [1985] 2 Lloyds R 103, CA) Any ambiguity or lack of clarity must be resolved against that party. Further the 'Standard Conditions' of bills of lading are not the subject of negotiation or amendment by the shipper; they are printed conditions which the shipper is required to accept ( ie a contract of adhesion); the wording is chosen by the issuer of the bill of lading Here the printed words relied upon by the shipowners are anything but clear, partly by reason of their context in the printed conditions and partly because, owing to misprints, they do not as printed make sense. As transferable documents of title needing to be understood internationally, merchants must be able to take them at face value. For the transferee, the relevant document is the bill of lading itself. The arguments advanced by Mr Gee on the first question now count strongly against him on the fourth.

However, for the purpose of the following discussion I will assume that the missing words in the second paragraph of clause 5 had been filled in with the corresponding words from the Himalaya clause in The New York Star . As regards the argument that the first part of the second paragraph of clause 5 should be construed as a contract not to sue, I do not consider that this is a possible construction of the wording.

The relevant authorities start with the Elder Dempster case. I do not need to add to what has been said about it by Lord Goff in The Pioneer Container ( sup .). It enabled the actual carrier to rely upon the same exceptions and limits as the 'contracting carrier'. The case which commences the modern sequence of cases to which reference must be made is Scruttons v Midland Silicones [1962] AC 446 . Its historical setting was that the High Court of Australia had recently in Wilson v Darling Island Stevedoring & Lighterage Co 95 CLR 43, another stevedore case, overruled two earlier New South Wales authorities which had enabled stevedores to rely upon the protective limits available to the shipowners under the bills of lading; Fullagar J, who delivered the leading judgment for the majority, acknowledging the principle of a bailment upon terms, distinguished the Elder Dempster decision. ([1956] 1 Lloyds at p.358); and in Adler v Dickson [1955] 1 QB 158 the Court of Appeal in England had held that an injured passenger could recover in full from crew members notwithstanding the exceptions in the passenger's contract with the shipowners. Scruttons were stevedores who had dropped the cargo owners' drum of chemicals whilst lowering it onto a lorry. The purpose of the action was to test whether the stevedores could rely upon the Hague Rules package limitation in the bill of lading. The stevedores were not bailees of the drum, "sub, bald or simple" (p.470). Accordingly no question of bailment on terms could arise and the stevedores could gain no assistance from the Elder Dempster case (p.470). The question was whether there was a principle of 'vicarious immunity' in English law or whether a contract between the tortfeasor and victim was necessary. The answer given (Lord Denning dissenting) was the latter: for the tortfeasor to be protected, he must have a legally enforceable contract with the victim which entitled him to enforce the exception or limitation against the victim. The House of Lords had an obvious distaste for unmeritorious manoeuvres to get round the Hague Rules exceeded only by their distaste for perverting the principled approach of English law in order to do so. To this end, Lord Reid suggested a principled solution. He said, at p.474,

The first case involving what is now called the 'Himalaya' clause (after the name of the cruise liner involved in Adler v Dickson ) to come before a court in London was New Zealand Shipping v Satterthwaite (The Eurymedon) [1975] AC 154 . New Zealand Shipping was in fact the parent of the shipowners. It had a longstanding course of dealing with them and acted as their stevedores and general agents in New Zealand. The relevant consignment (machinery) had been damaged in the course of discharge from the carrying ship as a result of the negligence of the stevedores' servants. It was not a bailment case. The cargo owners did not commence proceedings within the one year period permitted by Article III rule 6 of the Hague Rules; if that period is allowed to expire the cause of action is extinguished. It is a substantive provision not a procedural time bar: it extinguishes the liability. ( The Aries [1977] 1 Lloyds R 334, HoL) The stevedores relied upon the 'Himalaya' clause. Lord Wilberforce delivering the majority opinion said (p.166) that the case was not a challenge to the decision in Midland Silicones and expressly quoted and applied what had been said by Lord Reid ( sup ). There was no problem with the terms of the bill of lading which clearly gave both the carrier and the stevedore the benefit of the Hague Rules time limit and included appropriate words of agency (Reid, No.1 and No.2) nor in view of the evidence about the relationship between the stevedores and the shipowners, with the question of authority (No.3). The problem for Lord Wilberforce was the question of consideration or mutuality. He did not argue with the necessity for this if there was to be an enforceable contract. He seems to have preferred the view that there was a unilateral contract or offer by the shipper which became mutual when accepted by performance by the stevedores. (pp.167-8)

I would also observe that a principal theme of the judgment was expressed by Lord Wilberforce in these terms:

The next case was The New York Star [1981] 1 WLR 138 . Its facts were held to be indistinguishable from those in The Eurymedon . This time the appeal was from the High Court of Australia. Again, the cargo owners had omitted to commence their proceedings within the 1 year time limit and sought to get round this by suing instead the stevedores whose servants' negligence had caused their loss. The stevedores relied upon the 'Himalaya' clause. The cargo owners, ultimately unsuccessfully, relied upon a number other arguments which are not relevant to the present case. There were a wide range of opinions expressed in the Australian courts. In the High Court a dissenting judgment was delivered by Barwick CJ. ([1979] 1 Lloyds 298) The Privy Council, through Lord Wilberforce, expressly approved and adopted what he had said. (pp.144 and 148) Again the critical question was Reid No.4. No.1 was not in dispute and No.2 and No.3 (actual agency and prior authority) had been proved as facts. Barwick CJ proceeded from the Reid formula, which he quoted (p.303) but he was not wholly happy with the solution given in The Eurymedon to the problem of finding mutuality and consideration and suggested his own solution. The whole of his judgment is of value but for present purposes the quotations must be more limited (pp 304-305, 308):

In the Privy Council, Lord Wilberforce did not feel it necessary to enlarge on any of these points. He described his earlier judgment in The Eurymedon as having decided in principle that the 'Himalaya' clause was capable of conferring on third parties "defences and immunities conferred by the bill of lading upon the carrier" as if they were parties to "the contract contained in or evidenced by the bill of lading". (p.143)

I have not overlooked the judgment of Lord Goff in The Mahkutai [1996] AC 650 PC in which he conducted a forward looking review of the cases to which I have already referred. The interest for present purposes is that it involved shipowners who were the actual carriers. But the shipowners were attempting to enforce an exclusive jurisdiction clause against the bill of lading holders and failed as a matter of construction both under the 'Himalaya' clause and the principle of bailment upon terms. In reaching this conclusion, Lord Goff repeated a statement found in one form or another in all the 'Himalaya' clause cases:

What emerges from these cases?

The first point is that the shipowners are the actual owners of the ship named on the front of the bill of lading, the Starsin, and the employers of her crew. They took possession of the goods and carried them. The consideration which they are giving to the shipper in order to make clause 5 enforceable between them - to "enact" the arrangement in the clause, to use Barwick CJ's language - is the actual carriage of the goods. That is the service - the "act or acts" - which they are performing for the cargo owners. They assumed the possession of the goods as carriers by way of sub-bailment and entered into a bailee-bailor relationship with the shipper (see Lord Pearson, sup ). The bill of lading previously contained an inchoate contract of carriage - or "arrangement" - capable of becoming an enforceable contract when the shipowners enter upon the carriage of the shipper's goods. The shipowners have thus entered into a contract of carriage with the shipper. Both Lord Wilberforce and Barwick CJ treated the resultant contract as a mutual one but the mutuality arises from the performance of the service by the one party, here the carrying of the goods by the shipowners, and the agreement by the other that, should he do so, he shall be entitled to the stated protective provisions. Barwick CJ excluded an analysis involving an exchange of promises. Thus the person (the shipowners) performing the service (the carriage of the goods) does not promise in advance that he will do so nor that he will continue to do so. The 'Barwick' contract is not a contract with executory obligations; it is not a contract for the carriage of the goods. But it is a mutual contract of carriage which lasts as long as the shipowners remain the bailees of the goods as the actual performing carriers. The element of consideration and mutuality which the shipowners render to the shipper under the contract is their entering into this bailor/carrier relationship with the cargo owners and performing the carriage. In the event, they carried the cargo from Malaysia to Europe. Further, the contract is one which is "covered" by a bill of lading. Thus both the relevant person, here the shipowners, and the contract come within the definitions of "contract of carriage" and "carrier" in Article I (b) and (a) of the Hague Rules. Likewise, a contract of carriage which lasts only so long as the carrier concerned remains in possession of the goods as a carrier conforms to the Hague Rules definition of "carriage of goods" as covering "the period from the time when the goods are loaded on to the time they are discharged from the ship": Article I (e). (See also Article VII and the standard "Period of Responsibility" clause in the bills of lading.) The scheme of the Hague Rules is that they deal with the terms of the carriage not the destination to which they should be carried.

It may be asked why this consequence was not referred to in Midland Silicones , The Eurymedon or The New York Star . The answer is that they were all stevedore cases, as was the Canadian case ITO v Miida Electronics 28 DLR (4th) 641 referred to by my noble and learned friend Lord Steyn. (It is relevant to note in passing that the observations of McIntyre J in that case were made in the context of a carefully constructed contract between the carrier and the stevedore, the loss of the goods by theft subsequent to discharge, ie outside the Hague Rules period, and a 'Himalaya' clause which gave the stevedore no greater immunities than were given to the carrier by the bill of lading; contentious questions of the construction of the bill of lading nevertheless arose and required several pages of the judgment to resolve.) The stevedore is not a carrier. A stevedoring contract is not a contract of carriage. The 'Barwick' contracts did not include any element of carriage. A fundamental peculiarity of the present case is that the shipowners were the actual carriers who 'enacted' the 'Barwick' contract by becoming sub-bailees and performing the carriage of the claimants' cargo. The shipowners have escaped from being the original contracting carriers by relying upon the doctrine of privity of contract and the way in which the bills of lading were signed. They have brought themselves back in as a contracting carrier by relying upon clause 5 in the bills of lading and the privity of contract which it expressly creates.

It is argued that the 'Himalaya' clause contract is "collateral" to the bill of lading contract and therefore is not to be affected by such considerations as the Hague Rules. Why the use of the epithet "collateral" should have this effect is not clear. It does not address or affect the essential question: what is the 'Barwick' contract? In so far as a 'Himalaya' clause may include additional stipulations as between the person issuing the bill of lading and the shipper such as jurisdiction clauses or covenants not to sue, it may well be correct to use the word "collateral". But even then the substance may have to be looked at not just the form: The Hollandia [1983] 1 AC 565. But, as regards the persons referred to in the clause, clause 5 says that it is, for the purposes of all the provisions of the clause, made on behalf of such persons and to that extent all such persons shall "be deemed to be parties to the contract contained in or evidenced by this bill of lading". As between those persons and the shipper the resultant contract is not 'collateral'; it is the contract. The purpose of the additional use of these express words is to procure that transferees of the bill of lading shall be bound as well as the shipper: see the final sentence of the quotes from Lord Reid ( sup ) and Lord Wilberforce ( sup ). Clause 5 deliberately makes the contract between such persons and the shipper part of the bill of lading contract so as to obtain the benefit of it against other persons besides the shipper. Were it not for the inclusion of these words in the clause the shipowners would not have been able to rely upon as against any of the claimants in this litigation.

Then it is said that the contract is a "contract of exemption"; it merely exempts the other person. This may be a valid observation where the completion of the 'Barwick' contract does not involve the assumption of any special relationship towards the cargo owner. But, when the completion of the 'Barwick' contract involves becoming the sub-bailee of the goods and the actual performing carrier, the 'Barwick' is a contract of carriage, albeit purportedly on terms of complete exemption. True the contract does not include promises or executory obligations and will come to an end as soon as the relationship ceases, but that does not prevent the contract from being a contract and, while it subsists, a contract of carriage. To deny that it is a contract of carriage is to ignore the fact that the service being provided (and which makes the contract enforceable between them) and its subject matter is the carriage of the goods by the shipowners for the goods owner.

This leads on to the next point: what on the true construction of the bill of lading are the exemptions given to the shipowners as actual performing carriers? Is it a total exemption as stated in the opening words of the second paragraph of clause 5 or something less? Clause 5 provides, comprehensively, that "every exemption limitation, condition and liberty herein contained and every right exemption from liability, defence and immunity of whatsoever nature applicable to the carrier or to which the carrier is entitled hereunder shall also be available and shall extend to protect" the shipowners. Therefore one turns to clause 2 of the bill of lading

Clause 5 of the bills of lading is headed "Immunities" and starts with a long paragraph of exclusions from liability for loss or damage arising or resulting from a long list of possible causes. This paragraph is probably without any validity or effect whatsoever. In so far as it seeks to claim exemptions not permitted under the Hague Rules, it is under clause 2 and Article III r.8 invalid and void. In so far as it, in one or two instances, seeks to rely upon causes which are already excluded under the Hague Rules, eg , inherent vice (Art IV r.2(m)), it is redundant. The presence of this paragraph in this form of bill of lading is a clear illustration that substantial portions of the bill of lading terms may be without legal effect and/or be contradicted by other parts of the same clause or by other clauses of the bill of lading.

It is following this somewhat unpromising start that the reader arrives at the words relied upon by the shipowners, still part of the same clause but in a second and less long paragraph. It too starts with a blanket and comprehensive exclusion of any liability whatsoever. It is upon this part of the clause that the shipowners rely as containing the exemption. It is drafted widely enough to cover a very wide range of persons (including "any person who performs work on behalf of the vessel ( sic ) on which the goods are carried"). This part may present no problem for any entity, such as a stevedoring company, which is not a carrier and has no contract of carriage with the shipper. But for the shipowners it is by reason of clause 2 as ineffective as the preceding paragraph. It is not until one gets to the next part of the paragraph which contains the words given effect to in The Eurymedon and The New York Star , that the conflict with clause 2 of the bill of lading is resolved.

It is relevant at this point to observe in passing that the bills of lading already contemplate that there may be more than one carrier. Two clauses deal with this situation, clauses 16 and 17, respectively headed "Forwarding and Transhipment" and "Pre-Carriage and On-Carriage". Both of these subject matters are commercially commonplace in the international carriage of goods. They show that any application of clause 5 to carriers who are not the person who issued the bill of lading must be approached with caution and may raise problems of the construction of the bill of lading as a whole. This is not the place to discuss the law about the impact of the Hague Rules, and their successors, on these situations save to note that they do not countenance any evasion of the responsibilities and liabilities under those Rules. Clause 17 can be used as an illustration.

The same type of points can be made on clause 16 "FORWARDING AND TRANSHIPMENT" which is a much longer and more complex clause. It deals with a situation which is very similar to that arising in the present case, carriage of the goods on a ship which is not owned by the party on whose behalf the bill of lading was issued. It expressly contemplates more than one carrier. Where it applies, it contradicts and must displace clause 5. The scheme which it introduces is an authorised re-bailment or sub-bailment to another carrier. This is what happened in the present case, both factually and legally. The only difference in the present case is that the carriage on the Starsin was not pursuant to a liberty to tranship. In commercial terms, this is not a difference of substance. The same point can be made in relation to clause 16.

Therefore, applying The Eurymedon and The New York Star , the shipowners' attempt to get round clause 2 of the bill of lading and the Hague Rules fails and accordingly their defence fails. It is reassuring that this conclusion also conforms to the judicial statements as to the purpose (or function) of the 'Himalaya' clause and to the outcome which would be arrived at in jurisdictions which are not circumscribed by contractual criteria or by common law concerns with privity and consideration. Nor would a different conclusion have been arrived at under the Hague-Visby or Hamburg Rules. Nor would it be different if the Contracts (Rights of Third Parties) Act 1999 had applied; it would still have been necessary see what exemptions this bill of lading on its proper construction, including clause 2, conferred upon the shipowners. I can take the further points more shortly. Some are contentious.

This, as recognised by Barwick CJ, is a potential problem. (It was also the subject of comment by Lord Simon of Glaisdale in his dissent in The Eurymedon [1975] AC at p.179.) It was not necessary to decide the point in either of the leading cases because in each the stevedores was able to rely upon the time limit and that sufficed to discharge them from all liability. But the problem is a real one: an agreement without legal content is not a legally enforceable contract. (See the well known dictum of Devlin J in Firestone v Vokins [1951] 1 Lloyds 32 at p.39.) Failure to take account of this problem in considering Lord Reid's fourth head, will lead to absurdities. (Factual circumstances giving rise to absurdities can also be postulated.) Like Barwick CJ, I would wish to give much closer consideration to the question before accepting the proposition that the 'Himalaya' contract can support a total absence of legal responsibility. However in the present case this point does not arise upon the correct construction of this bill of lading and, in a case such as the present, it may suffice that the third party actually assumes the possession of the claimant's goods together with the resultant bailee-bailor relationship and, indeed, joins in a common maritime adventure with the legal consequences that that entails, eg general average.

The relevant elements were proved as facts in The Eurymedon and The New York Star . But the present case is not such a case. There was already a contract between the person issuing the bill of lading and the 'sub-contractor'. It was the timecharter. This made express provision for what any bill of lading contract involving the shipowners was to contain. This did not include any form of Himalaya clause. The bill of lading was to contain a Hague Rules clause paramount ("as attached"). It did not contain any authority to stipulate for a complete immunity. It should be remembered that one of the purposes of a clause paramount is to avoid a situation where the carrier by claiming too much loses even that which he could have claimed: see Article III r.8, inf . Thus the clause paramount provided: "If any term of this bill of lading be repugnant to any extent to any legislation by this clause incorporated, such term shall be void to that extent but no further." The preservation of this situation is of real value to an actual carrier and may well be a condition of its P&I cover. As previously explained the sub-bailment was on these terms. Also, the Hague Rules are not without other benefits to the carrier, for example the liability of the shipper in connection with the shipment of dangerous goods.

How then is it said that a carriage performed on Hague Rules terms can, after all performance has been completed, become a contract performed on terms of complete immunity? (It cannot satisfy the 'Barwick' analysis which requires a contract or 'arrangement' already made at the time of the issue of the bill of lading.) It is said: by the carrier later asserting a complete immunity in answer to subsequent litigation. I venture to suggest that this is both legal and commercial nonsense. It arises in the present case because the party invoking clause 5 is the actual carrier who had an express agreement with the timecharterer as to the terms upon which he was to carry the goods and what was to be put in the bills of lading and is then attempting to use the clause retrospectively to evade the Hague Rules. It shows that to accede to the shipowners' case involves the acceptance of very far reaching and anomalous propositions which would find no resonance in other jurisdictions. It is another reason why in my opinion the shipowners' defence must fail in this case.

During the discussion in argument of the references to "the ship" or "the vessel" in the bill of lading clauses and in the Hague Rules, particularly in Article III r.8, Mr Gee made various submissions in order to fit the words to his argument. It is correct that, say, in a charterparty forming part of a chain of charterparties in respect of the same voyage, each person chartering out the vessel assumes the role of her owner and, on a contractual incorporation of the Hague Rules, the legal obligations and liabilities of a carrier and that the same applies the other way round to the charterer in each charterparty, say, for freight. Each is acting as a principal and the charterparties should be construed accordingly. (Chains of commodity contracts are another example.) It is also correct to say that in some circumstances considerable verbal manipulation is permissible to give effect to what the court considers is the clear intention of the parties. The most striking example of this is Adamastos Shipping v Anglo-Saxon Petroleum [1959] AC 133 where a clause paramount simply attached to a consecutive voyage party was held to apply to delays during the ballast legs. But the fact that Mr Gee's arguments do require some verbal manipulation of the bill of lading wording to make them fit does disclose that questions of construction and reconciliation do arise as I have made clear in relation to clause 5 on the one hand and clauses 16 and 17 on the other.

Article III r.8 provides

Mr Gee also cited and placed particular emphasis upon three New South Wales cases. The first was Gadsden v Australian Coastal Commission [1977] 1 NSWLR 575, CA. The defendants were the owners of a ship which carried the plaintiffs' goods from Port Kembla in New South Wales to Kwinana Western Australia. During the course of the carriage the goods were damaged as a result of the negligence of the defendant's servants. The ship was at the time under timecharter. The bills of lading were issued and signed on behalf of the timecharterers not the defendants. They did not contain any form of 'Himalaya' clause. The action was started after one year had expired. The Court of Appeal held that nevertheless the action should succeed. The defendants put their case on two grounds. The first was that, relying upon the Elder Dempster case, they could rely upon a clause in the bill of lading incorporating the time limit. The second was that they had in some way a statutory right to the defence. Moffitt P and Samuels JA (Hutley JA agreeing) summarily rejected the first defence as involving an unjustified extension of Elder Dempster and being inconsistent with the Wilson v Darling Island case, giving no other reasons save for adopting whatever the trial judge had said . Why this was thought to be so is not clear and is contrary to the judgment of the Privy Council in The Pioneer Container and, on my reading, with Fullagar J's judgment in Wilson v Darling Island . ( Gilchrist Watts & Sanderson was referred to in argument but not in the judgments nor is there anywhere any reference to the terms of the timecharter.) This case provides no answer to Lord Goff's reasoning. Once the first defence was rejected, the second defence did not stand a chance. The defendant was a stranger to the bill of lading and any contract with the plaintiffs. The legislation in force in Australia at that time was still the Sea Carriage of Goods Act 1924 and its effect was only to incorporate the Hague Rules compulsorily into bills of lading contracts of carriage. Since the plaintiffs had no contract, they had no gateway into the statute. They sought to treat the statute as providing them with a freestanding defence under Article III r.6 because that rule included the words "and the ship". This was a hopeless argument because it did not address the relevant flaw in their case but it provided Samuels JA with the opportunity to discuss the nature of the action in rem in the course of which Samuels JA said at p.585:

The other two cases can be taken more shortly. The Gadsden case was applied in Sydney Cooke v Hapag-Lloyd [1980] 2 NSWLR 587, Yeldham J. The proposition for which the judge cited it (p.595) was "that a 'carrier' in order to be subject to the obligations and subject to the rights and immunities conferred in the Hague Rules must be a party to the contract of carriage covered by the bill of lading or similar document of title in so far as the latter relates to the carriage of goods by sea". This raises no question not already discussed (and would appear to support my conclusion). The bill of lading was a consolidation combined bill of lading. It was issued by the first defendant Hapag-Lloyd whose ship carried the container from Hamburg to Sydney. At Sydney the container was discharged and delivered to the depot of the second defendants where the container was unstuffed by the second defendants and the plaintiffs' consignment set aside to await collection. Owing to the negligence of the second defendants' servants the goods were damaged. The plaintiffs sued the second defendants as the party liable, joining Hapag-Lloyd as first defendants. Hapag-Lloyd applied for the action to be stayed on the ground that it contravened the covenant in the bill of lading against suing sub-contractors. The judge granted the stay. The arguments of the plaintiffs were essentially all based on construing the elaborate bill of lading clauses which treated the contract as being divided up into separate stages. Although the anti-suit clause had some similarities to parts of a Himalaya clause, the judge said that the clauses under consideration in The New York Star were quite unlike those in the bill of lading before him. (p.595)

Finally there is the question of the lack of clarity and certainty in these bills of lading. As printed, clause 5 did not have the effect for which the shipowners contend. This was in effect the main reason why the judge and the Court of Appeal rejected the shipowners' case on clause 5. I have not found it necessary to decide this appeal on this ground. But it is a real point none the less. If a party chooses to put an ineffective exclusion clause in his contract, it is not for the courts to fashion one which will give him what he ought to have asked for. The opposite party is entitled to rely upon has been written and is not obliged to go away and research what might have been written. This applies a fortiori to a transferable documentary contract. As is again illustrated by Mr Gee's Australian cases, there are quite a number of different kinds and versions of 'Himalaya' type clauses. In the present bills of lading, the reader does not even have the assistance of a relevant heading to the clause. If the shipowners' contentions in the present case at least recognised the judicially stated function of the Eurymedon/New York Star clause, the position might be capable of some moderation. But they do not. They run directly contrary to that function.

For these reasons I would allow the appeal of the shipowners in respect of all consignments except those of Makros Hout BV. The judgment in favour of Makros Hout BV should stand.

I shall confine my remarks to what appear to me to be the two issues of general importance: (i) the identity of the parties to the bills of lading; and (ii) the validity of the contractual exemption from liability which the Himalaya Clause in the bills of lading purports to afford to the owner or demise charterer of the ship on which the goods were laden.

The Identity of the Carrier.

As the trial Judge observed, it would strike any one unfamiliar with maritime law, as I am, as quite extraordinary that there should have grown up an immense body of decided cases devoted to this issue. For my own part, I regard it as not only extraordinary but lamentable.

The identity of the parties to a contract is fundamental. It is not simply a term or condition of the contract. It goes to the very existence of the contract itself. If it is uncertain, there is no contract. Like the nature and amount of the consideration and the intention to create legal relations it is a question of fact and may be established by evidence. Such evidence is admissible even where the contract is in writing, at least so long as it does not contradict its express terms, and possibly even where it does: see Young v Schuler (1883) 11 QBD 651 Chitty on Contracts 28th edn p 633. But bills of lading are transferable documents of title, and the claimants are holders of the bills by endorsement. Consequently the evidence must be found within the four corners of the bills themselves.

Where a contract is contained in a signed and written document, the process of ascertaining the identity of the parties and the capacity in which they entered into the contract must begin with the signatures and any accompanying statement which describes the capacity in which the persons who appended their signatures did so. This may require interpretation, and to this extent the process may without inaccuracy be described as a process of construction. But it is not of the same order as the process of construing the detailed terms and conditions of the contract. These describe the incidents of the contract and the nature and extent of the parties' obligations to each other. But the identity of the parties themselves is not an incident of the contract. Where a signature is accompanied by a description of the capacity in which the signatory has appended his signature the description is not a term or condition of the contract. It is part of the signature and so part of the factual evidence of the identity of the party which is undertaking contractual liabilities under the contract.

The bills of lading in the present case were on pre-prepared printed forms. Printed on the front of each form in large letters are the words "Liner Bill of Lading, the name "Continental Pacific Shipping" and the group's logo. The bills of lading are thus on Continental Pacific Shipping's own printed forms.

The forms themselves were prepared and intended for use as owners' bills. On the front of each bill was printed a testimonium for signature by the Master. The Master would, of course, be the natural person to sign on behalf of the owner or demise charterer of the ship. The detailed terms in miniscule print unreadable by the naked eye on the back of each form were also those appropriate for an owner's bill. Clauses 33 (Identity of Carrier) and 35 (Demise Clause), though not identified as such by any heading or marginal note, stated in terms that the party which was undertaking the contract of carriage was the ship owner or demise charterer as the case might be. Since the printed forms were unmistakably Continental Pacific Shipping's forms and were unambiguously intended for use as owners' bills, the inference must be that Continental Pacific Shipping had vessels of their own and that the forms were prepared for the carriage of goods on such vessels, whether or not they were subject to a time charter in favour of another.

The front of each bill of lading contained, as one would expect, boxes to be completed with the particulars of the intended voyage. There were boxes for the names and addresses of the shipper and consignee, the ports of loading and discharge, particulars of the goods, and the name of the vessel on which they were to be laden. These were filled in typescript. The name of the vessel was given as " M.V. Starsin".

As completed prior to signature, therefore, the bills of lading appeared to be owners' bills for the carriage of goods on Continental Pacific Shipping's vessel M.V. Starsin prepared for signature by the Master of the vessel.

Continental Pacific Shipping, however, were not the owners or demise charterers of this vessel. They were only time charterers. Moreover none of the bills was signed by or on behalf of the Master. They were all signed by Port Agents "As Agents for Continental Shipping (the Carrier)", or with variants thereof. All the contractual obligations of carriage set out in the small print on the back of each bill were undertaken by "the Carrier", and Clause 1(c) identified "the Carrier" as "the party on whose behalf this Bill of Lading has been signed", ie Continental Pacific Shipping.

What we have, therefore, are charterers' bills, that is to say bills signed for and on behalf of the charterer as the party undertaking the contract of carriage, but on printed forms intended for use as owners' bills. To my mind the problem to which this gives rise presents no difficulty. It is not appropriate to embark on the unrewarding task of attempting to reconcile the irreconcilable. The charterers must simply be taken to have used the wrong form, probably because it was the form that they or their agents were accustomed to use. Their agents signed the bills on their behalf as contracting parties, and any of the detailed terms and provisions on the back of the printed forms which are inconsistent with their use as charterers' bills must be modified accordingly or treated as not applicable: see The Okehampton [1913] P 173 at p 180 per Hamilton LJ.

This conclusion can be reached by a number of different routes, all of which yield the same result. It is a well established canon of construction that, where there is inconsistency between the printed terms of a standard form and the terms which the parties have themselves written into the document, the latter should prevail: see Robertson v French (1803) 4 East 130 at p 136 where Lord Ellenborough distinguished between the written words as "the immediate language and terms selected by the parties themselves" and the printed words of the form as "a general formula adapted equally to their case and that of all other contracting parties upon similar occasions and subjects." The passage has been approved in your Lordships' House in Glynn v Margetson [1893] AC 351 (and by Scrutton LJ in Re An Arbitration between L Sutro & Co and Heilbut, Symons & Co [1917] 2 KB 348 at pp 361-2.

This principle is applicable even where the inconsistent provisions are of equal importance and the printed form is appropriate to the particular case as well as to the general. How much more must primacy be given to the written words where they describe the main intent and object of the particular contract. In Glynn v Margetson the House refused to permit a printed clause in a standard form of bill of lading to defeat the object and intent of the contract as derived from the words written into the contract by the parties themselves. The printed form was applicable to many voyages; the written words were specially agreed upon in relation to the particular voyage.

This principle is not peculiar to English law. Article 305b of the German Civil Code, for example, provides

In every contract some terms are fundamental. In the case of a contract for the sale of land, such terms are the parties, the property, and the price. In a contract of carriage, they are the parties, the goods, the vessel, and the ports of loading and discharge. Such terms are both necessary and sufficient. They describe the main object and intention of the contract. If any of them are not agreed, there is no contract. If they are agreed, all else is detail.

Some at least of such terms may, of course, be varied in particular circumstances by the detailed terms and conditions of the contract. These may, for example, provide for a reduction or increase in the consideration in a specified event, or for transhipment or deviation, so long as the variations do not defeat the main object of the contract. But the identity of the contracting parties is of a different order altogether. I do not think that this is capable of variation by the detailed terms and conditions of the contract. How can provisions which purport to prescribe the incidents of a contract between A and B convert it into a contract between B and C?

A further consideration which it is relevant to bear in mind is that of market practice. I need no persuasion that businessmen expect the identity of the carrier, together with other variables which describe the object of the particular voyage, such as the vessel, the goods, and the ports of loading and discharge, to be found on the face of the bill of lading and not tucked away among the standard terms and conditions printed on the back. Since 1994 the practice of the market has been adopted by the ICC Uniform Customs and Practice for Documentary Credits, Article 23 of which indicates that banks do not in practice examine the contents of the terms and conditions of carriage on the reverse of a bill of lading. That of course is a matter between the bank and its customer, not between the shipper or consignee and the carrier; but against such a commercial background it would create an unacceptable trap to allow the detailed conditions on the back of a bill of lading to prevail over an unequivocal statement of the identity of the carrier on the face of the bill.

All these considerations yield the same conclusion, so that there is no need to choose between them. Cumulatively they indicate not only that the Canadian Pacific Shipping was the carrier, but that it used the wrong form. I would make this the basis of my own decision.

As a fall-back position the Respondents submit that the owner of the ship and the time charterer are both liable as contracting parties. The submission must be rejected. All the provisions of the bills, both back and front, indicate the existence of a single carrier, though they do not agree upon its identity. Two inconsistent provisions cannot be reconciled by adopting a construction which is consistent with neither. The carrier must be one or the other; it cannot be both.

Accordingly, and in respectful agreement with the trial Judge and with Rix LJ, I would hold these bills to be charterers' bills.

The Himalaya Clause.

I agree with all your Lordships on the construction of Clause 5 of the bills of lading (the Himalaya Clause). The clause does not make grammatical sense as it stands, and it is obvious that words have been omitted. The Court must, therefore, supply the omission by implying at least the minimum necessary for the clause to make grammatical sense. This is what all the judges below did. But the authorities show that in a proper case the Court will go further. Where it can see, not only that words have been omitted, but what those words are, then it is its duty to supply them. It is not necessary that the Court should be certain precisely what words have been omitted; it is sufficient that it knows their gist. The process is one of construction, not rectification; this is evident from the fact that the Court of Chancery not infrequently supplied omissions in wills at a time when it had no jurisdiction to rectify them.

The Respondents submit that it is inappropriate to adopt such a liberal approach in the case of an exemption clause. More should not be read into such a clause, they say, than is absolutely necessary to make sense of it, particularly where to do more would relieve a party from any obligation at all: see Tor Line AB v Alltrans Group of Canada Ltd [1984] 1 WLR 48 at pp 58-9. I cannot accept this as a proposition of universal application. I agree, of course, that an exemption clause must be strictly construed and not given a wider application than can fairly be derived from the words used. But the argument in the present case is not concerned with the scope of the exemption, but with the presence or absence of the mechanism of agency which is critical to the efficacy of the clause. The principle which the Respondents invoke has no application in such a case.

It is obvious that the clause is not an original work of legal draftsmanship but is taken from a precedent. Several versions of the Clause are in circulation, and it is impossible to identify the particular precedent from which the defective clause in the present case was taken. But they all employ the same mechanism of agency to give legal efficacy to the clause; they all do so by identical or nearly identical words; and they all incorporate the mechanism at precisely that part of the present clause where words have been omitted. In my opinion this is a clear case where the Court can and should supply the missing words.

I also agree with all your Lordships that the opening words of the clause are words of exemption; they cannot be construed as a covenant not to sue. They are the very antithesis of such a covenant, which presupposes the continued existence of liability, albeit a liability which the covenantor undertakes not to enforce. The present clause by contrast operates to prevent any liability from arising in the first place.

It is well established by the authorities that the Himalaya Clause has the effect of bringing into being a separate or collateral contract between the cargo owner and a third party, usually an independent contractor such as a stevedore, under which the third party enjoys exemption from liability to the cargo owner. They also establish that the contract is a unilateral or "if" contract by which the third party undertakes no obligation to the cargo owner of any kind, but the cargo owner promises that if the third party does anything in the course of its employment which damages the cargo it will have the benefit of the protective provisions of the clause: see The Eurymedon [1975] AC 154 at p 168 per Lord Wilberforce; The New York Star [1981] 1 WLR 138 The Makhutai [1996] AC 650 .

Such a contract is a promise for an act, not a promise for a promise. If in the course of its employment the third party performs an act in relation to the goods, which it is under no obligation to the cargo owner to perform, it will at the one and same time bring the contract with the cargo owner into existence and supply the consideration for the cargo owner's promise of exemption from liability. In The New York Star Lord Wilberforce, at p 144, expressly approved the analysis of the clause which was adopted by Barwick CJ in the same case in the High Court of Australia (though contrary to the general usage he preferred to regard it as a bilateral contract): see [1979] 1 Ll Rep 298 at p 305 where he said:

As my noble and learned friend Lord Hobhouse of Woodborough has observed, the present case goes further than any previous decision. In all previous cases, it seems, the third party has either been a stevedore or some person other than the owner or demise charterer of the ship ( The Eurymedon ; The New York Star ; Sidney Cooke Ltd v Hapag Lloyd Aktiengesellschaft [1980] NSWLR 587; Chapman Marine Pty Ltd v Wilhelmson Lines A/ S [1999] FCA 178; or has relied on the clause not to claim exemption from liability but merely to claim the benefit of a time limit or jurisdiction clause ( The Eurymedon; The New York Star ; The Makhutai ; The Pioneer Container; J Gadsden Pty Ltd v Australian Coastal Shipping Commission [1977] NSWLR 587). In the present case the third party is the owner of the ship on which the goods were carried and relies on the clause to claim a complete immunity from liability at the suit of the cargo owner.

But the wording of the clause is plain and unambiguous. It provides ( inter alia ") that

In many of the cases the function of a Himalaya Clause has been described in limited terms: ie to make available to the third party the same exemptions and limitations as are available to the carrier. This was sufficient for the purpose of those cases, and I do not read them as limiting the purpose for which the Clause may legitimately be employed in other cases.

I am satisfied, therefore, that, as a matter of construction, Clause 5 of these bills of lading creates a collateral contract between the cargo owner and the owner or demise charterer of the ship which exempts the latter from all liability to the cargo owner if it undertakes the actual carriage of the goods. This brings me to the critical question, which is whether the exemption of the owner or demise charterer of the ship on which the goods were carried (as distinct from a mere stevedore for example) from liability for loss or damage to the goods is contrary to the Hague Rules, to which the bills of lading were expressly made subject. This is a very difficult question, on which I have changed my mind more than once.

Article III Rule 8 of the Hague Rules invalidates any condition of a contract of carriage which purports to relieve "the carrier or the ship" from liability for loss or damage to the goods. The Hague Rules apply only to contracts of carriage by sea covered by a bill of lading or similar document and to which the party which claims relief from liability is a party: see Article I(b) and J Gadsden Pty Ltd v Australian Coastal Shipping Commission . The question, therefore, is whether the contract between the cargo owner and the owner or demise charterer of the ship created by the Himalaya Clause is a contract for the carriage of goods by sea within the meaning of the Hague Rules.

In my opinion it is not. A contract of carriage of goods is a contract by which one party, the carrier, undertakes a contractual obligation to carry the goods. Each of the bills of lading contained or evidenced such a contract between the charterer and the cargo owner. It could not legitimately relieve the charterer from liability for loss or damage to the goods, and it did not purport to do so.

But the collateral contract between the cargo owner and the owner or demise charterer of the ship contained in the Himalaya Clause did not oblige the owner or demise charterer of the ship to carry the goods. It merely exempted it from liability to the cargo owner if it did so. During the voyage the owner or demise charterer of the ship would not be carrying the goods pursuant to any contract with the cargo owner. It would merely be providing consideration for a promise on the part of the cargo owner that it would not be liable for any damage which might be caused to the goods.

Such a contract cannot properly be characterised as a contract of carriage. It is rather a contract of exemption which is ancillary or collateral to other contractual arrangements (the time charter and the bill of lading) which were necessary to achieve the carriage of the goods on the chosen vessel. This is the conclusion reached by the Australian cases to which I have referred, and with which I respectfully agree. In The Eurymedon Lord Simon of Glaisdale, who dissented in the result, said (at p 179) that he confessed to difficulty in grasping the very concept of a contract which consisted only of an exemption clause. The majority agreed that the content of the contract was as he described it, but did not consider that this gave rise to any conceptual difficulty.

The question, of course, is not whether the contract created by the Himalaya Clause would be characterised as a contract of carriage as a matter of English law, but whether it is such a contract within the meaning of the Hague Rules. The terms of Article III Rules 1 and 2 indicate that it is not. They import into a contract of carriage of goods by sea positive obligations on the part of the carrier (including the owner or demise charterer of the ship) which enters into a contract of carriage with the cargo owner: (see Article I(a)). These include obligations relating to the loading, handling, and stowage of the goods. It would be remarkable if a contract entered into for the sole purpose of granting a party exemption from liability had the paradoxical effect of subjecting it to liabilities to which it would not otherwise be subject.

Accordingly, I am satisfied that the Himalaya Clause is not itself a contract of carriage of goods by sea, and that merely by taking the benefit of such a Clause the owner or demise charterer of the ship does not become a party to a contract of carriage and so a carrier within the meaning of Article I(a) of the Hague Rules.

But the matter does not rest there. Clause 5 of the bills of lading in the present case provides inter alia that the third party who takes the benefit of the provisions of the Clause shall "to this extent" be deemed to be a party to the contract contained in or evidenced by the bill of lading. The purpose of these words is to bind subsequent holders of the bill of lading by the limitations on their rights of suit imposed by the Himalaya Clause: see Section 2(1) of the Carriage of Goods by Sea Act 1992. The words "to this extent" are words of limitation. They have the effect of deeming the third party to be a party to the bill of lading to the extent of obtaining the benefit of the protective provisions of the Clause but no further. They do not subject the third party to any of the contractual obligations of the carrier, including the obligation to carry the goods, contained in or evidenced by the bill of lading.

What are the consequences? As I understand the argument, they can be summarised in three propositions:

This is an attractive and elegant solution to the problem. It enables Article III Rule 8 to serve its purpose, which is to ensure that the security of the ship is available to the cargo owner notwithstanding any contractual provision to the contrary, without producing the paradoxical consequences of incorporating Article III Rules 1 and 2. But is it correct?

My initial view was that it is not. My difficulty was with proposition (ii). It would normally be an abuse of language to describe a person as being a party to a contract of carriage because he is a party to a contract which is such a contract only because someone else undertakes obligations of carriage thereunder. It must be a contract under which he undertakes such obligations himself.

But on further reflection I have changed my mind. Article III Rule 8 of the Hague Rules invalidates any provision in a contract of carriage of goods by sea which relieves " the carrier or the ship " from liability for loss or damage to the cargo. As I have pointed out, the word "carrier" includes the owner or demise charterer of the ship which has entered into a contract of carriage. Unless the words "or the ship" are tautologous, therefore, they must be intended to cover the case where the owner or demise charterer of the ship has not entered into a contract of carriage. Thus the Rule invalidates a provision contained in a contract of carriage covered by a bill of lading to which the owner or demise charterer of the ship is a party and which purports to relieve the owner or demise charterer of the ship from liability for loss or damage to the cargo even though it has not itself entered into a contract of carriage. The only way in which effect can be given to such a provision is to square the circle and accept that the owner or demise charterer of the ship can become a party to a contract of carriage covered by a bill of lading even though it is a contract under which it does not itself undertake any obligations of carriage.