England and Wales Court of Appeal (Civil Division) Decisions

IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE QUEEN'S BENCH
(Mr Justice Morison)

		Royal Courts of Justice Strand, London, WC2A 2LL Thursday 6 December 2001

LORD JUSTICE SIMON BROWN:

1. If an independent expert departs from his instructions in a material respect his determination is not binding. What for these purposes is a material respect? That is the critical question before us. It arises on appeal from Morison J's order in the Commercial court on 2 May 2001 dismissing the defendant's application for summary judgment under CPR Part 24.2.
2. The circumstances in which the point arises are not in dispute and can be briefly told. By a contract made in August 1999 the appellants (the Sellers) sold a cargo of 25,000 plus metric tonnes of gasoil to the respondents (the Buyers) FOB Antwerp.
3. The contract provided, amongst other things:

"4. Product/Quality Gasoil meeting the following guaranteed specifications:

Test Limit Method ASTM

Density at 15degC 0.876kg/l max D1298"

That test was followed by 20 other specified tests.

"10. Quantity/Quality

Quantity and quality to be determined by a mutually agreed independent inspector at the loading installation, in the manner customary at such installation. Such determination shall be final and binding for both parties save fraud or manifest error. Inspector to be appointed by Seller. Costs to be shared equally between Buyer and Seller."

4. On 20 August 1999 some 34,000 mt of cargo was loaded on board the vessel m.t. "ROBIN" at the installation of the Belgian Refining Corporation BV at Antwerp. Caleb Brett were the mutually agreed independent inspectors instructed to carry out the determination under clause 10. By their report dated 26 August 1999 they recorded that the cargo had been sampled and analysed and that its density at 15 ºC was 0.8750 kg/l (i.e. within the contractual specification of 0.876), using test method D4052 (rather than the specified method D1298).
5. In this action, brought on 17 August 2000, the Buyers claim some US$250,000 from the Sellers. They do so on the basis that they on-sold the cargo to the Lebanese Ministry of Oil on similar terms as to quality but that on arrival and testing the density of the oil was found to exceed the contractual maximum. They dispute the Inspectors' determination. This, they say, is not final and binding because the wrong testing method was used.
6. The Sellers contend that the Inspectors' determination is binding and conclusive and that the Buyers' claim is accordingly bound to fail. Hence their application for its summary dismissal and, following Morison J's rejection of their argument below, this appeal.
7. Before coming to the arguments I should record certain further matters of agreement. First, that test method D4052 is more modern and accurate than D1298, having a margin of error of .0001% as opposed to 0.0007%. Second, that had the Inspectors used method D1298, they would inevitably still have found the density test satisfied in respect of the actual samples tested. Third, that method D4052 as opposed to D1298 is that customarily used at Antwerp.

Was there a departure from instructions?

8. Before turning to what I have already described as the critical issue, it is necessary first to dispose of a preliminary argument: the Sellers' contention that, on the proper construction of the contract, the Inspectors were not required to use test method D1298 and accordingly are not to be regarded as having departed from their instructions, materially or otherwise.
9. Mr Nolan's argument in this regard runs essentially as follows. Clause 4 is a specification clause pure and simple requiring no more than that the cargo should have a density up to the specified limit. Clearly one way of determining as a matter of objective fact whether that specification was met was by using the prescribed method. But it was not the only way. Another method could be used if, as here, it would still be possible to demonstrate that the specification would have been met even had the prescribed method been used. The only provision in the contract expressly providing for how the independent inspector was to determine the quality of the cargo was clause 10 which provided for this to be done "in the manner customary at such installation". The word "manner" in that context encompasses all aspects of the determination of cargo quality including testing methods. Mr Nolan, indeed, went so far as to submit that the Inspectors were required to use customary method irrespective of whether it was better or worse than the method prescribed by clause 4.
10. For my part I would reject these arguments. Clause 4 does not require that the cargo must be a cargo the specification of which is such that if it were to be tested by the specified method D1298 it would meet the density specification but that it need not be so tested. On the contrary, it provides that this test method must be used. And, indeed, why else would the test method be specified? Why should the parties care whether the cargo is theoretically capable of satisfying a given test unless that particular test is to be used? Mr Nolan submits that test D1298 is specified merely as a standard or bench mark test and that any better test would suffice. If his reasoning is sound, however, a less accurate test (provided always it was "customary" at the installation) would also suffice so long as it could be shown that it would have produced the same result – as, indeed, it would have done here (if one postulates the specification of test D4052 and the use of D1298).
11. There is this further objection to Mr Nolan's argument: in order to determine whether the Inspectors' determination is binding, the parties (or the court) must go through the exercise of comparing the result of the actual test used with the result of a hypothetical test using the specified method. True it is that the parties here have sensibly agreed what the outcome of such an exercise would be. That, however, will not always be so and often the exercise would be problematic. It would certainly be uncommercial.
12. In short, I share the view expressed by the judge below that clause 10 is not to be read on its own. Clause 4 identifies both the standard and the method for assessing whether the standard has been reached. What was required was a test conducted by the stipulated method and none other. Clause 10 deals with the "manner" of carrying out the required tests. This would, of course, include the method where that was not otherwise specified (as was so in the case of some tests under clause 4). But the "manner" would include a host of other matters too, not least the sampling procedures to be followed. These are in fact what the Buyers complain of and will seek to criticise if the Inspectors' determination is set aside. The Inspectors' evidence that the samples tested met the contractual specification will be well-nigh unassailable. Not so, however, the suitability or sufficiency of the samples taken.

Departure in a material respect

13. Having concluded that the Inspectors departed from their instructions, I come then to the more difficult question of whether this departure was in a material respect. At an early stage of his submissions, and certainly in his written argument, Mr Nolan suggested that there is a distinction to be drawn between on the one hand a departure which is de minimis or trivial and on the other hand one which is immaterial. Ultimately, however, he came to accept that all these expressions are synonymous. The critical question is what in this context they entail. The Sellers' argument in a nutshell is that a departure is not material unless it could have had an effect upon the ultimate result, i.e. could have affected the Inspectors' determination that the cargo was of the specified quality, and here it is conceded it could not, and indeed did not.
14. The Buyers' contrary argument is that once, as here, a clear departure from instructions is established, the court should not be drawn into the exercise of determining what if any effect it might have had on the end result. Mr Goldstone submits that only the most trifling departure should properly be characterised as immaterial, where, for example, a valuer arrives late for an appointment. Failing that, Mr Goldstone submits that a departure should not be regarded as immaterial unless the court is satisfied that it could have made no conceivable difference not merely to the outcome of the determination itself but also in any other respect, as for example to a third party's readiness to accept, or pay against, the determination. There are many possible reasons, unconnected with the achievement of an accurate physical result, for choosing a particular test. It could have related to the on-sale contract. It could have related to a letter of credit. It could simply have been the test with which the Buyers were most familiar and confident. It is not for the court to speculate. It is sufficient that the parties have agreed it.
15. Before turning to the authorities most closely in point, it is convenient first to recognise two principles which inevitably touch on the issue. The first, and that on which understandably Mr Nolan places reliance, is to be found in Cairns LJ's judgment in Toepfer v Continental Grain Co. [1074] 1 Lloyds Reports 11, 14:

"When parties enter into a contract on terms that the certificate of some independent person is to be binding as between them, it is important that the Court should not lightly relieve one of them from being bound by a certificate which was honestly obtained and not vitiated by fraud or fundamental mistake on the part of the certifier. When, for instance, as in this case, the certificate called for by the contract is one relating to the quality of goods sold, the business purpose is to avoid disputes about quality, and that purpose is defeated unless it is made difficult for a party to go behind a valid certificate. "

16. The second, clearly countervailing, principle is surely this: inspectors should be astute to comply with their instructions and, if they depart from them, there should not then be much scope for dispute and litigation as to whether their determination is nevertheless binding. In short, the interests of finality cut both ways although, of course, one bears in mind that if a determination is set aside the underlying dispute is left unresolved.
17. Coming then to the authorities I propose to take as my starting point the first instance decision of Jones(M) v Jones (RR) [1971] 1 WLR 840. The court there was concerned with a valuation produced by a valuer who departed from his instructions in two respects: first he valued shares on a break-up basis whereas he was instructed to use a going-concern basis; secondly he valued certain machinery himself whereas his instructions required him to have it valued by an expert valuer of his choice. Ungoed-Thomas J at p.854 turned to address:

"… the defendant's contention that an error in principle does not vitiate a valuation unless it is shown by the person relying on it that it also results in a materially different valuation, both in the part of the valuation subject to error and in the overall valuation."

18. The two cases with which he was principally concerned were Dean v Prince [1954] Ch 409 and Frank H. Wright (Constructions) Limited v Frodoor Limited [1967] 1 WLR 506. The valuers in Dean v Prince, as in Jones v Jones itself, had valued shares on a break-up rather than a going-concern basis but, unlike in Jones v Jones, that had not been contrary to their instructions; rather their instructions had simply been to determine a "fair value". The Court of Appeal unanimously upheld the valuation although with some reluctance. Lord Evershed MR at p.426 said this:

"… I have felt compelled to the conclusion that although, as I think, Mr Jenkinson [the valuer] erred in principle in treating himself as bound on accountancy principles to regard only the break-up value, Mrs Dean has nevertheless failed sufficiently to establish that a consideration of the values of the plant 'in situ' or otherwise would produce, in all the circumstances, a figure of value materially different from that at which Mr Jenkinson arrived."

Denning LJ at p.427 stated the principle thus:

"… if the courts are satisfied that the valuation was made under a mistake, they will hold it not to be binding on the parties."

19. Frank H Wright (Constructions) Limited v Frodoor concerned an error on the face of the certificate and decided that an immaterial error does not vitiate a valuation. But Roskill J said at p.529: "If this error had been material, it would have been enough to vitiate the whole of the certificate, small as it might be and regrettable as the consequences might be." He then added with regard to the error in the case before him: "This error is not material because it does not affect the result."
20. In Jones v Jones at p.855, Ungoed-Thomas J said of Roskill J's judgment:

"He was not there dealing with any question of burden of proof; nor – which is crucial – was he dealing, and nor indeed was Evershed MR dealing [in Dean v Prince] with a valuation made in a manner contrary to directions binding upon the valuer as to the manner or method of valuation."

21. A little later, at p.856, Ungoed-Thomas J continued:

"… I do not conclude that there is any requirement of general application that where a valuation is made on an erroneous principle, yet the valuation nevertheless stands unless it is also shown that a valuation on the right principle would produce a materially different figure from the figure of the valuation that he made. (This would incidentally place on the objector the onus, not only of proving that the selected expert has acted on the wrong principle, but of incurring what might be the very heavy burden and expense of a completely new valuation, which itself might not be accepted as conclusive between the parties and merely leading to yet another valuation.) The authorities thus to my mind establish that if a valuation is erroneous in principle, it is vitiated and cannot be relied upon even though it is not established that the valuation figure is wrong."

I take "erroneous in principle" in that final sentence to refer to a departure from instructions.

22. I come next to this court's decision in Jones v Sherwood Services Limited plc [1992] 1 WLR 277 upholding an expert's report on the basis that the experts had done precisely what they had been instructed to do. In the course of a judgment which considered a large number of authorities, Dillon LJ noted first, the development of the law in the mid-1970's establishing that an expert could be liable for damages if he had acted negligently in giving his certificate (see the House of Lords' decisions in Sutcliffe v Thackrah [1974] AC 727 and Arenson v Arenson [1977] AC 405); secondly, that this required reconsideration of the principle that a certificate could be vitiated for mistake (how could an expert be liable for a negligent mistake in giving a certificate if the effect of that mistake was that the certificate was not binding on the parties?); and thirdly, that accordingly, in Campbell v Edwards [1976] 1 WLR 403 and Baber v Kenwood Manufacturing Co. Limited [1978] 1 Lloyds Reports 175, this court "look[ed] at the question of setting aside certificates of experts on grounds of mistake afresh in the light of the principle that the expert or valuer can be sued for negligence". Lord Denning MR had said in Campbell v Edwards at p.407:

"It is simply the law of contract. If two persons agree that the price of property should be fixed by a valuer on whom they agree, and he gives that valuation honestly and in good faith, they are bound by it. Even if he has made a mistake they are still bound by it. The reason is because they have agreed to be bound by it."

23. In Jones v Sherwood Dillon LJ observed:

"Plainly Lord Denning came to change his views between 1954 [in Dean v Prince] and 1976 [in Campbell v Edwards]. … We also therefore are free to look at the matter afresh on principle, and are not bound by the law as stated by common consensus in Dean v Prince. … [at p.286]

On principle, the first step must be to see what the parties have agreed to remit to the expert, this being, as Lord Denning MR said in Campbell v Edwards … a matter of contract. The next step must be to see what the nature of the mistake was, if there is evidence to show that. If the mistake made was that the expert departed from his instructions in a material respect – e.g. if he valued the wrong number of shares or valued shares in the wrong company, or if, as in Jones v Jones … the expert had valued machinery himself whereas his instructions were to employ an expert valuer of his choice to do that – either party would be able to say that the certificate was not binding because the expert had not done what he was appointed to do

… Because Coopers [the valuers] did precisely what they were instructed to do, the plaintiffs cannot challenge their determination of the amount of the sales.

There is another line of reasoning which points in the same direction, though I do not found my judgment on it. [at p.287]

…

Therefore, if there was any 'mistake' over the contracts subject to ratification, it is a wholly immaterial mistake with no practical effect at all: see the judgment of Roskill J in Frank H. Wright (Constructions) Limited v Frodoor …, where he refers to an error being 'material' only if it materially affects the ultimate result." [at p.288]

24. The final authority to which I come on this part of the case is Shell UK v Enterprise Oil [1999] 2 AllER (Comm) 87. Put shortly, the experts there, in breach of contract, had used the wrong computer programme to map an oil field. Lloyd J held in a reserved judgment following a six day hearing that that departure from the experts' instructions vitiated their final decision. In the section of the judgment headed "The law as regards materiality of a breach", Lloyd J discussed and quoted from, principally, Frank H. Wright (Constructions) Limited v Frodoor, Jones v Jones and Jones v Sherwood. The first two of those cases, he noted, had not actually been cited to him although, of course, both were referred to in Jones v Sherwood. The relevant paragraphs in Lloyd J's judgment are these:

"97. Materiality must, of course, also be considered in the light of the particular contract. In this case, the contract lays down the procedures for submissions and documents passing between the parties and the expert according to a carefully regulated procedure and timetable for what is likely to be, at best, quite a lengthy process but one which is to be pursued at considerable speed. In that context, it seems to me that the test of materiality has to be capable of being applied, at the latest, at the moment when the error first comes to light, which is likely to be in the first formal document produced by the expert after the error has occurred, or on a request for clarification by one or another party soon thereafter. The ultimate effect of the error on the parties' position may not be known, or even capable of being forecast with any accuracy, especially if it arises at an early stage. Both parties agree that it cannot be necessary or possible to wait until the outcome or effect of the error is known, which may be a long way down the line, before being able to decide whether the error is sufficiently material to vitiate the expert's act. It is therefore a question of assessing materiality by reference not to whether it actually affects the ultimate result, but according to its potential effect on the result and, perhaps even more importantly, on the process, including the ability of the parties to manage and deal with the procedure in accordance with the contract.

98. I should also say that, if the expert has committed a material breach of instructions, then as a matter of law the relevant act is not binding on any of the parties, leaving aside of course the effect of their subsequent acts. It is not a point on which the court has a discretion whether or not to allow the expert's act to stand. I do not consider that Lightman J intended to suggest that there was such a discretion when summarising the law in British Shipbuilders v VSEL Consortium plc [1997] 1 Lloyd's Rep 105 at 109, even though he said that the court 'may' set the decision aside. That he did not mean to indicate that it was a discretionary issue appears in any event from the next following sentence. The relevant passage is:

'(3) If the expert in … his determination fails to comply with any conditions which the agreement requires him to comply with in making his determination, the court may intervene and set his decision aside. Such a determination by the expert as a matter of construction of the agreement is not a determination which the parties agreed should affect the rights and duties of the parties, and the Court will say so.'"

As Lloyd J had already noted, Lightman J in the British Shipbuilders' case was for this purpose drawing essentially upon Jones v Sherwood.

25. Mr Nolan submits, as stated, that a departure is not material unless it affected, or at least could have affected, the outcome of the determination. In so submitting he relies upon a number of passages from the judgments I have discussed, most notably perhaps Lord Evershed's reference in Dean v Prince to "a figure of value materially different from that at which [the valuer] arrived"; Roskill J's characterisation of the error in Frank H Wright (Constructions)Limited v Frodoor as "not material because it does not affect the result"; Dillon LJ's other "line of reasoning" in Jones v Sherwood referring to "a wholly immaterial mistake with no practical effect at all," with express reference back to Roskill J's judgment; and Lloyd J's reference in paragraph 97 of his judgment in Shell UK v Enterprise Oil to not waiting to see "whether the error is sufficiently material to vitiate the experts' act" so that materiality is to be assessed "according to its potential rather than its actual effect upon the ultimate result." All these passages, Mr Nolan argues, support the view that the essential touchstone of materiality is whether or not the departure might have affected the outcome of the determination. If demonstrably it could not have done so, or if, as here, that is conceded, the determination should stand.
26. Strongly and skilfully though the argument was advanced, for my part I cannot accept it. Rather I see the position as follows:

i) A mistake is one thing; a departure from instructions quite another. A mistake is made when an expert goes wrong in the course of carrying out his instructions. The difference between that and an expert not carrying out his instructions is obvious.

ii) Under the old law a mistake would vitiate the expert's determination if it could be shown that it affected the result. That was the concept of material mistake established in Dean v Prince and Frank H. Wright (Constructions) Limited v Frodoor. Not so, however, with regard to a departure from instructions – see Ungoed-Thomas J's judgment in Jones v Jones cited in paragraph 21 above.

iii) Under the modern law the position is the same as it was with regard to a departure from instructions, different with regard to mistakes. As Lord Denning explained in Campbell v Edwards, if an expert makes a mistake whilst carrying out his instructions, the parties are nevertheless bound by it for the very good reason that they have agreed to be bound by it. Where, however, the expert departs from his instructions, the position is very different: in those circumstances the parties have not agreed to be bound.

iv) The test of materiality devised for identifying vitiating mistakes does not carry across to the quite separate field of departures from instructions. This seems to me so both as a matter of principle and of authority. The position is as stated in Jones v Jones and in Dillon LJ's judgment in Jones v Sherwood at p.287 (quoted in paragraph 23 above) where he illustrates the principle by reference to Jones v Jones.

v) Dean v Prince and Frank H. Wright (Constructions) Limited v Frodoor – although on any view rightly decided – should no longer be regarded as authoritative with regard to experts' mistakes. That for the most part was made clear in Jones v Sherwood. The contrary is not to be inferred from the dictum in Dillon LJ's judgment at p.288 (the other line of reasoning on which he did not found his judgment) referring back to Frank H. Wright (Constructions) Limited v Frodoor. It is time that Dean v Prince and Frank H. Wright (Constructions)Limited v Frodoor received their quietus.

vi) Once a material departure from instructions is established, the court is not concerned with its effect on the result. The position is accurately stated in paragraph 98 of Lloyd J's judgment in Shell UK v Enterprise Oil: the determination in those circumstances is simply not binding on the parties. Given that a material departure vitiates the determination whether or not it affects the result, it could hardly be the effect on the result which determines the materiality of the departure in the first place. Rather I would hold any departure to be material unless it can truly be characterised as trivial or de minimis in the sense of it being obvious that it could make no possible difference to either party.

27. I should add that each side before us sought to draw some comfort from Lloyd J's judgment in Shell UK v Enterprise Oil: Mr Nolan, as indicated, because of the references in paragraph 97 to the error (clearly there a departure from instructions) potentially if not actually affecting the ultimate result; Mr Goldstone because the judge's reference to the effect of the departure on the "process" shows at least that he had broader considerations in mind than merely the result of the test itself. It is perhaps a pity that Frank H. Wright (Constructions) Limited v Frodoor and Jones v Jones were not explored in argument before the judge: the suspicion arises that he may not have appreciated that the former, as pointed out in the latter, was not concerned with a departure from instructions. This may explain what to my mind was Lloyd J's excessive concern with the effect of the expert's breach of instructions on the process if not the result.
28. I therefore agree with what Morison J said in paragraph 15 of his judgment below:

"The word 'material' as used by Dillon LJ in Jones v Sherwood at p.287 ('if the mistake made was that the expert departed from his instructions in a material respect') is capable of meaning that either the mistake must be more than trivial or that the mistake must have made a material difference to the result. If it were the latter, then the protection afforded by a final certificate clause would become less effective since the court would have to try the facts to see if the error made a material difference. Since the whole purpose of the final determination provision is designed to avoid such a trial, I cannot think that the Court of Appeal meant more by the word material than 'not de minimis.' As Knox J pointed out in Nikko Hotels v MEPC [1991] 2 EGLR 103, it is not the business of the court to weigh the importance of a stipulation in a contract. If the requirement to use the method specified in clause 4 was contractual that is an end of the matter, whether or not the court thinks it was important or would have made a difference. The assumption the court makes, on normal principles of construction, is that the parties have chosen to define their rights and obligations according to their own needs, whatever they may have been. In this case, with an on-sale, one can envisage the possibility that the Buyers wanted the old fashioned test for reasons connected with it. However, that is speculation. The position is, I think, that it is no business of the court to enquire why the Buyers asked for this particular test. That was what was agreed … "

29. In short, the Inspectors here were required to determine the quality of the cargo using test method D1298. They did not do so and the parties had not agreed to be bound by a determination as to quality by any other method. The determination is accordingly not binding.

Manifest error

30. Although that conclusion is sufficient to dispose of the appeal, I would touch briefly on the alternative basis for decision relied upon by the Buyers, the reference in clause 10 to "manifest error". Morison J below went no further than to say that he was "inclined to the view that there was a manifest error here, due to the wrong test being used".
31. Morison J had previously considered the meaning of "manifest error" in Conoco (UK) Ltd v Phillips Petroleum (unreported, 19 August 1996) where, following dicta in earlier cases, he held that manifest error referred to: "oversights and blunders so obvious as to admit of no difference of opinion".
32. The question then arising is whether it is relevant to consider whether the error is one that affected the result. Considering that question in Conoco v Phillips, Morison J said this:

"… it seems to me that there is no room for any debate as to whether the oversight or blunder would or would not have made any material difference to the result. If it could be shown that there was a manifest error then in my judgment that would be an end of the case. If fraud was shown, I cannot accept that it would be open to debate as to whether the fraud did or did not affect the result; so also would manifest error."

33. I confess to some difficulty with this approach. Fraud, of course, would vitiate the determination irrespective of whether it affected the result: "Fraud or collusion unravels everything" (per Lord Denning in Campbell v Edwards). The exception for "manifest error", however, seems to me of a rather different character and to be designed essentially to fill the gap in the law created by the development to which I have already referred: the overthrow of the Dean v Prince principle of setting aside determinations for mistake. Nowadays, if parties wish to contract on the basis that they will not be held to mistakes made by the expert in the course of carrying out his instructions, they must needs include a term like this with regard to manifest error. But if they do, is it then really to be said that provided only the mistake is obvious, the determination will be avoided irrespective of whether it could affect the outcome? In this context I am inclined to think not. Take the very error committed in Frank H. Wright (Constructions) Limited v Frodoor, the erroneous inclusion of a 'not' in the report. I do not think that that ought properly to be regarded as a "manifest error". Rather I would extend the 'definition' of manifest errors as follows: "oversights and blunders so obvious and obviously capable of affecting the determination as to admit of no difference of opinion". (emphasis added).
34. If, of course, the error consists of a departure from instructions, then, assuming I am right in my earlier conclusion, it will never be necessary to ask whether in addition that error amounts to a "manifest error": it will vitiate the determination in any event. If, however, I am wrong in my earlier conclusion – if, in short, the Inspectors' use of the wrong test method here ought properly to be regarded as an immaterial departure from their instructions – I would not conclude that it nevertheless constituted a manifest error such as to entitle the Buyers to set aside the determination on that alternative basis.
35. That, however, is by the way. For the reasons given earlier, I would hold the determination not to be binding because of the Inspectors' material departure from their instructions, and accordingly dismiss this appeal.

LORD JUSTICE TUCKEY :

36. Like Simon Brown and Dyson LJJ, I agree that this appeal should be dismissed. There is however, an apparent difference between them as to how to define a material departure from instructions about which I think I should express an opinion so that anyone looking for guidance from our judgment will know what the majority view about this is.
37. My choice is between "any departure is material unless it can truly be characterised as trivial or de minimis in the sense of it being obvious that it could make no possible difference to either party" (the first test – Simon Brown LJ at para. 26 (vi)) or a departure which "the parties would reasonably have regarded …. as sufficient to invalidate the determination" (the alternative test – Dyson LJ at para. 47).
38. I do not think the legal route which enables the test to be formulated matters although I prefer the implied term analysis. If, of course, there is an express term that is another matter, as Dyson LJ points out in para. 44. But the first test is just as capable of being an implied term as one which derives from the principle of de minimis non curat lex.
39. I have to say I prefer the first test because I think it is more certain and direct. There may in fact be no difference between the two tests in practice, but an enquiry as to what the parties would reasonably have regarded as material appears at least to be less certain and direct than the enquiry required by the first test. Whichever test is applied the court would obviously need to consider the subject matter and express terms of the contract, the nature of the departure and any other relevant facts.
40. I see no reason why application of the first test will only result in departures of form or procedure being excused. If, in the rare case postulated by Dyson LJ in para. 48, it is clear beyond argument that a departure of some other kind could not affect the result however the result is characterised, then it is obvious that it could make no possible difference to either party. If the passage from the judge's judgment referred to by Simon Brown LJ in para. 28 precludes this possibility, I agree with Dyson LJ's reservations about it, but I do not read it in this way. Rather I think it expounds the first test.

LORD JUSTICE DYSON:

41. I agree with Simon Brown LJ that this appeal must be dismissed. I only wish to add a few observations on the "departure in a material respect" point. The leading authority is Jones v Sherwood Computer Services [1992] 1 WLR 277. Much of the debate in the present case has centred on the passage in the judgment of Dillon LJ:

"If the mistake made was that the expert departed from his instructions in a material respect- eg if he valued the wrong number of shares or valued the shares in the wrong company….either party would be able to say that the certificate was not binding because the expert had not done what he was appointed to do".

The question is: what did he mean by "in a material respect"?

42. As he pointed out, the first step is to see what the parties have agreed to remit to the expert, this being a matter of contract. In the present case, as Simon Brown LJ has explained, the parties agreed to refer not merely the question of whether the density of the oil was below 0.876kg/l at 15 deg C, but whether it was below that level when tested by ASTM method D1298.
43. In saying that the determination would not be binding if the expert departed from his instructions in a material respect, I believe that Dillon LJ was referring to what the parties must be taken to have agreed. Since they did not agree this expressly, they could only have done so impliedly. In my view, a term should be implied in an expert determination clause such as clause 10 in the present case to the effect that the determination will not be binding if the expert departs from his instructions in a material respect. Such a term is reasonable and represents the obvious, but unexpressed, intention of the parties.
44. I prefer to analyse the problem in contractual terms than by reference to the general principle that the law does not concern itself with trifles: de minimis non curat lex. After all, it is possible, at least in theory, for the parties to a contract to specify the instructions to be followed by the expert in the utmost detail, and to stipulate expressly that any departure whatsoever, however trivial, is to be regarded as a material departure.
45. But where the parties have not defined what departures are to be regarded as material, the question remains: what is the yardstick by which materiality is to be judged? The judge contrasted a departure that was more than trivial or de minimis with one that must have made a material difference to the "result". He said that any departure would be material unless it was trivial. The fact that it made no difference to the result could not save the determination.
46. I have some misgivings about the suggested antithesis between on the one hand a departure that is de minimis and on the other hand a departure that must have affected the result. As I have already explained, the question is one of contract: what did the parties intend? It is not a question of the application of the general legal principle that de minimis non curat lex. Moreover, one has to be careful to define what one means by the "result" in this context. In the present case, if the result is that the sample, when tested, showed a density of less than 0.876kg/l at 15deg C, then it is common ground that the departure did not affect the result. If, however, the result is that the sample, when tested in accordance with D1298, showed a density of less than 0.876kg/l at 15 deg C, then the departure did affect the result.
47. So what is the test by which materiality is to be judged? Surely it is simply whether the parties would reasonably have regarded the departure as sufficient to invalidate the determination. At one end of the spectrum will be departures of form or procedure which could have no bearing on the substance of the determination. Unless the parties have so agreed expressly or by necessary implication, it will be a rare case in which a court would hold that they have impliedly agreed that such a departure would invalidate a determination. At the other extreme will be significant departures of substance. Dillon LJ gave two good examples of these in the passage that I have quoted.
48. In deciding what the parties must be taken to have regarded as material, the court will take into account the subject-matter and express terms of the contract and all the relevant circumstances. I would not rule out the possibility that there may be cases in which a departure from instructions would reasonably be regarded by the parties as immaterial because, although the expert's mistake is not one of form or procedure, it is clear beyond argument that the departure could not affect the result, however the result is characterised. Such a mistake is immaterial because the parties could not reasonably consider it to be material. But such cases are likely to be rare because it will not often be possible to say with confidence that (a) the departure could not have affected the result, and (b) it was unreasonable for a party to regard a departure that could not have affected the result as material. Nevertheless, such cases may occur. It follows that I do not entirely accept the analysis made by the judge in the passage that has been quoted by Simon Brown LJ.
49. Turning to the present case, I am in no doubt that the fact that the use of the wrong method cannot have affected the "result" does not save the determination. As Mr Goldstone has pointed out, it cannot be assumed that the choice of a particular method of testing is a consequence solely of the parties' desire to achieve an accurate result. There may be different reasons which have been dictated by the terms of other related contracts and/or letters of credit. The possibility that there may be such other reasons is by no means far-fetched in the context of a commercial contract which is likely to be one of a chain. These are not matters about which the court can or should speculate. The starting point is that if the parties have agreed that a determination using method A is to be binding, then a determination using method B will not be binding because the parties have not agreed that it will be. It follows that the determination is not binding and the appeal must be dismissed.
Order: Appeal dismissed; appellant to pay respondent's costs in the sum of £12,448.40 by consent; application for permission to appeal to the House of Lords refused. (Order does not form part of the approved judgment)