Paula Lee Ltd v Robert Zehil Ltd QUEEN'S BENCH DIVISION [1983] 2 All ER 390 HEARING-DATES: 21 APRIL, 27 MAY 1982 27 May 1982 CATCHWORDS: Contract — Damages for breach — Alternative obligations — Defendant free to choose method of fulfilling contractual obligation — Basis for assessment of damages — Damages to be assessed according to method of performance least unfavourable to defendant — Defendant's freedom of choice limited to reasonable methods of performance. HEADNOTE: The plaintiffs, who were dress manufacturers, appointed the defendants to act as their sole distributors for the sale of their range of garments in a territory consisting of certain Middle Eastern countries.  Under the terms of the agreement the defendants undertook to purchase not less than 16,000 garments each season and by cl 7 of the agreement it was agreed that the defendants were to have complete discretion on the marketing and selling policy to be adopted in the territory.  Thus, provided the defendants ordered not less that 16,000 garments per season, they were entitled to sell or market the plaintiffs' products to whoever and on such terms as they wished within the territory.  The agreement did not make any stipulation as to the sizes or styles of the garments comprising the minimum obligation under the agreement.  The defendants prematurely terminated the agreement while it still had two seasons to run, and in an action brought by the plaintiffs for breach of contract the defendants were held to have wrongfully repudiated the contract.  On the issue of the quantum of damages there arose the question of how the plaintiffs' loss attributable to the premature termination of the agreement was to be computed.  The defendants contended that since cl 7 of the agreement gave them complete freedom of marketing in the territory they were entitled under the agreement to fulfil the whole order by choosing 32,000 of the plaintiffs' cheapest garments and therefore the damages should be the plaintiffs' loss of profits on 32,000 of their cheapest garments.  The plaintiffs contended that the appropriate yardstick by which the damages were to be measured was the average price of all garments in the plaintiffs' range which were suitable for sale in the territory since the defendants would have chosen 32,000 garments of varying styles, prices and quality from throughout the range. Held -- (1) Where damages were required to be calculated for breach of contract in a situation where the defendant could fulfill his part of the contract by performing his obligation by alternative methods and had a freedom of choice which method to use, damages were to be assessed by reference to that method of performance which was least unfavourable to the defendant.  However, in making that assessment the contract was to be read subject to an implied term in the contract that the defendant's freedom of choice was limited to those methods of performance which could be regarded as reasonable in all the circumstances (see p 394 c d and p 396 d, post) Abrahams v Herbert Reiach Ltd [1922]1 KB 477 applied Thomas v Clarke and Todd (1818) 2 Stark 450, Capper v Forster (1837) 3 Bing NC 938 and Cockburn v Alexander (1848) 6 CB 791 considered.   (2) Although the agreement gave the defendants a wide discretion as to the sale of the garments in regard to volume, price and outlets, it could not have been in the contemplation of the parties when they made the agreement that the defendants would order 32,000 garments of the same size, style and colour, since a one-garment range would have been unsaleable and would have alienated the defendants' wholesalers, deprived the plaintiffs of a realistic presence in the territory and killed the market.  Accordingly, the purchase of 32,000 of the plaintiffs' cheapest garments was not, in the circumstances, a reasonable method by which the defendants could fulfill their obligation under the contract.  Instead, the agreement was to be construed as being subject to an implied term that the 32,000 garments would be selected in a reasonable manner, and the damages were to be assessed in terms of that reasonable selection which yielded the lowest and therefore least unfavourable price to the defendants (see p 396 g to p 397 c, post). NOTES: Notes For the measure of damages in contract where the defendant has contracted to do one of a number of things and does neither, see 12 Halsbury's Law (4th edn) para 1177, and for cases on the subject, see 17 Digest (Reissue) 105, 127--129. CASES-REF-TO: Abrahams v Herbert Reiach Ltd [1922] 1 KB 477, CA. Bold v Brough Nicholson & Hall Ltd [1963] 3 All ER 849, [1964] 1 WLR 201. British Westinghouse Electric and Manufacturing Co Ltd v Underground Electric Rlys Co of London Ltd [1912] AC 673, [1911--13] All ER Rep 63, HL. Capper v Forster (1837) 3 Bing NC 938, 132 ER 672. Cockburn v Alexander (1848) 6 CB 791, 136 ER 1459. Lavarack v Woods of Colchester Ltd [1966] 3 All ER 683, [1967] 1 QB 278, [1966] 3 WLR 706, CA. Maredelanto Compania Naviera SA v Bergbau-Handel GmbH, The Mihalis Angelos [1970] 3 All ER 125, [1971] 1 QB 164, [1970] 3 WLR 601, CA. Thomas v Clarke and Todd (1818) 2 Stark 450, 171 ER 702. INTRODUCTION: Action By a writ issued on 26 October 1979 the plaintiffs, Paula Lee Ltd, claimed against the defendants, Robert Zehil & Co Ltd, damages for wrongful repudiation of a contract made between the parties on 21 August 1978 and amended by a supplemental agreement dated 26 June 1979 whereby the defendants agreed to act as the sole distributors of the plaintiffs' garments in Kuwait and Saudi Arbia.  The action was first tried on the issue of liability of the parties and on 21 December 1981 Mustill J held that the defendants were liable for anticipatory breach of contract for having wrongfully repudiated the contract with the plaintiffs.  The matter was adjourned on the issue of quantum of damages.  The case is reported only on the question of damages.  The facts are set out in the judgment. COUNSEL: Peter Millett QC and David di Mambro for the plaintiffs. Leslie Joseph QC and Victor Levene for the defendants. JUDGMENT-READ: Cur adv vult 27 May.  The following judgment was delivered. PANEL: MUSTILL J JUDGMENTBY-1: MUSTILL J JUDGMENT-1: MUSTILL J.  This is the second stage of an action brought by Paula Lee Ltd against Robert Zehil & Co Ltd.  The dispute raises issues of principle on both liability and quantum.  It was therefore arranged that the first part of the trial should deal with the question of liability, and with two issues relating to the quantum of damage, leaving the remainder of the dispute for decision at a second hearing if the plaintiff succeeded in establishing liability.  In the event, shortage of time made it impossible to give adequate consideration to the second issue on quantum, and I therefore confined my judgment to the question of liability and to one of the points raised on quantum.  At a second hearing full argument was addressed on this outstanding question and I now state my decision on it, by way of guidance to those who will conduct the matter when the facts relating to the issues of damages are explored.   The history of the dispute is set out fully in my previous judgment, and it is sufficient for present purposes to recall that the matter arose under a contract whereby the plaintiffs appointed the defendants to act as their sole distributors for the sale of their dresses in certain Middle Eastern countries, that the agreement was prematurely terminated in the manner which I have held to be a wrongful repudiation by the defendants, and that there remained, according to the interpretation which I have placed on the contract, two seasons of the arrangement left to run, at the time when it was repudiated.   It is convenient to set out again certain clauses material to the present stage of the argument.  The agreement, as subsequently amended, provided: '1. The Company hereby appoints the Distributor to be its sole and exclusive Distributor for the sale of its merchandise (hereinafter called ''the merchandise'') in the Countries Kuwait and Saudi Arabia (hereinafter called ''the territory'') subject to the terms and conditions hereinafter set out . . .   3. The Company hereby undertakes that all its exports to other Middle Eastern Countries including North Africa (excluding the territory) will be made directly by the Company and not through any agent and the Company shall refuse to supply any agent if it knows or ought to know that that agent is supplying customers in other Middle Eastern Countries including North Africa   4. (a) The Company will prepare two ranges per year of Girls dresses of a type traditionally made by it--the first range available in March/April and the second available, in September/October of each year  (b) The Distributor undertakes to purchase or procure the purchase of not less than Sixteen thousand (16,000) garments each Season from the Company for the Territory by the end of the sixth week of the date that the Distributor is informed by the Company in writing or by Telex that the Company's complete range of garments is available for showing . . .   7. The Distributor shall have the sole discretion on marketing and selling policy in the territory so that provided the Distributor is ordering not less than the said Sixteen thousand (16,000) garments per season the Distributor shall be able to sell or market the Company's products to whom they shall desire within the territory and upon such terms and conditions as the Distributor shall solely determine . . .' The question now for consideration is this.  Given that the contract provides only for the purchase by the defendants of at least 16,000 dresses, without saying anything about the sizes and styles of the dresses composing the minimum obligation, what assumption should be made as to the nature of the hypothetical purchase when computing the plaintiffs' loss attributable to the premature termination of the agreement?  For example, should it be assumed that the whole quantity would have been composed of the cheapest dresses?  If this is so, the plaintiffs will recover no damages for they sold sufficient of the more expensive models by way of mitigation to overtop the loss of profit on the larger quantity of the cheaper kind.  Or should the putative sales be related to an average, and, if so, what kind of average, price for the range as a whole?  Or should a forecast be made of the dresses which would have been sold if the contract had gone ahead?   There are two reasons why this is a difficult question to answer.  First, the scheme of the contract is not fully worked out in the document, and it is not easy to complete the scheme by implication.  Second, the principles on which damages are to be calculated in situations where the defendant has some freedom of choice as to the manner of performance are not so clearly established that they can easily be applied to the novel situation now arising.   At first sight, it must seem a surprising assertion that the theory of damages in relation to alternative obligations is still open to doubt.  In commercial disputes, damages are so often calculated in terms of the minimum quantity of goods to be delivered or cargo to be shipped that this mode of assessment has become a matter of routine, and the principles rarely have to be considered.  The principles are, however, less clear cut than might be thought.  For example, a dictum of Maule J which is very often cited from Cockburn v Alexander (1848) 6 CB 791 at 814, 136 ER 1459 at 1468--1469 to the effect that when assessing damages the presumed mode of performance is that which is least profitable to the plaintiff and the least burdensome to the defendant is a possible source of difficulty.  Unless this is only an elaborate way of saying that the performance is assumed to be that which will yield the least award, the two halves of the rule are capable of leading to different answers, for the position of the defendant might be better overall if he chose a mode of performance which involved a greater than minimum liability in damages but avoided losses in other directions.  Again, the foundation of the doctrine is very often sought in the statement of Scrutton LJ in Abrahams v Herbert Reiach Ltd [1922] 1 KB 477 at 482 that a defendant is not liable in damages for not doing that which he is not obliged to do.  This dictum looks to minimum performance, not to minimum recovery (although the two will often be the same), nor to minimum detriment for the defendant.  Furthermore, the principle is expressed solely in terms of the defendant's obligations, considered in the abstract.  Yet there are circumstances where the court has paid regard to evidence of what the defendant would actually have done if the contract had gone ahead: see, for example, Bold v Brough Nicholson & Hall Ltd [1963] 3 All ER 849, [1964] 1 WLR 201, Maredelanto Compania Naviera SA v Bergbau-Handel GmbH, The Mihalis Angelos [1970] 3 All ER 125, [1971] 1 QB 164 and perhaps also Abrahams v Herbert Reiach Ltd itself.   I believe that some at least of the difficulties which arise in this branch of the law can be minimised if it is kept in mind that inquiry always involves a comparison between the plaintiff 's actual position in face of the breach, and the position which he would have occupied if the contract had been performed.  This must involve an identification of the promise, followed by a valuation of its promised worth to the promisee.  Each part of the inquiry may involve considering a choice which would have been open to the promisor.  Thus:   1. The promisor may have a right of election which fixes the content of his obligation.  This can take more than one form.  It may, for example, give the vendor the option to deliver 1,500 to 2,000 tons of goods.  Here, in accordance with the dictum of Scrutton LJ, the damages are assessed on the basis that the quantity delivered would have been 1,500 tons, for the seller could not have been compelled to deliver more than this amount.  Where the obligation is to deliver at A or B, the dictum does not always work so well, for it cannot necessarily be said that A represents more of an obligation than B.  So here the presumption is explained in different terms, by looking for the obligation which would have been least detrimental to perform.   2. Even where the obligation has been fixed in advance, or determined by election, the value of it to the promisee, and hence the amount which he has lost through non-performance, may be determined by contingencies.  Sometimes, it is possible to be sure what would have happened if the contract had been performed, and (if so) this finding is used to estimate the worth of the promise.  More often, the fact that the repudiation has prevented the time for performance from arising means that the best that can be done is to make an estimate of the likelihood that the contingent event would occur, and adjust the damages accordingly.  For this purpose it makes no difference that the contingency is one which is under the control of the defendant.  Although it is not legitimate to look at what the promisor would have done, but only what he could have done, when identifying the promise, the position is different when, as in The Mihalis Angelos and Bold v Brough Nicholson & Hall Ltd the inquiry concerns the valuation of the promise itself: see especially per Diplock LJ in Lavarack v Woods of Colchester Ltd [1966] 3 All ER 683 at 691, [1967] 1 QB 278 at 295--296.   There is one further distinction which must be mentioned, namely that which exists between (a) an obligation expressed in terms of a range of alternatives from which the promisor may choose and (b) a single obligation expressed in an indefinite way.  A duty of the latter kind may often be construed as an obligation to act reasonably, and the damages will be assessed on the basis of what would have been reasonable.  That this distinction does exist cannot, I think, be disputed, and it presents no serious theoretical difficulty when it is possible to say that there is one reasonable mode of performance, and one alone.  But what of the case where there is more than one reasonable method, or a whole range of reasonable methods, shading into one another?  One possible view is that the court should try to forecast how the defendant would have performed but for the repudiation.  In my opinion this approach is inconsistent with principle, since the defendant may in the event have done no more than was necessary to qualify as reasonable, and to assess damages on any other basis would be to penalise him for failing to do something which he was not obliged to do.  The answer must, in my judgment, be that the court is to look at the range of reasonable methods, and select the one which is least unfavourable to the defendant, bearing in mind, of course, that in deciding what methods qualify as reasonable the question must be approached with the interests of both parties in mind.  This is, I believe, the way to account not only for the decision in v Herbert Reiach Ltd, but also for the divergencies of approach which might seem to exist between the various judgments, and within the individual judgments, delivered in that case.   I now turn to the present contract.  One thing at least is clear, that so far as quantity is concerned this is not a case of optional obligations.  There was a single obligation, to purchase 16,000 dresses, coupled with a liberty to purchase more if the defendants so wished.  The problem is to determine the content of the obligation in the absence of any provision as to the ratios in which the dresses were to be ordered.   Before discussing this question it is convenient to deal with an argument addressed by the plaintiffs to the effect that the damages should be computed by applying to the shortfall the average price of all of those dresses which were of a style and size suitable for sale in the territory.  There are, I believe, two ways in which this argument could be advanced.   The first is by the orthodox route of construing the contract first, and then measuring the damages in terms of the obligation thus arrived at.  If the argument proceeds in this way, it is in my judgment bound to fail.  Assuming for the moment that the plaintiffs are contending for a simple arithmetical average, the only obligation which would produce the desired award of damages would be one to order the minimum 16,000 dresses with every size and style in exactly the same quantities across the entire range.  This would be commercial nonsense, and could not be a tenable interpretation of the contract.  The proposition is not made more attractive by adopting some kind of weighted average, because there is nothing in the contract to fix the method of weighting.  It was suggested that a weighting might be arrived at by looking at past experience but this would not be practicable for the first season, and in any event to suggest that in each season the defendants were obliged to buy the minimum quantity of dresses in the same proportions as in the previous season, no matter how different the conditions (or for that matter the dresses) might be, is not a commercially feasible solution.  One further proposition, namely that the damages should be assessed by reference to the plaintiffs' own experience when selling in mitigation, plainly cannot be justified by means of any argument which takes the identification of the obligation as its starting point.   The plaintiffs do, however, have another basis for their argument in support of an average price.  This is founded on a small group of nineteenth century decisions relating to the computation of damages for breach with an obligation to load under a charterparty.  When approaching these rather difficult cases it is necessary to bear in mind that there are three questions which are distinct, although they appear to have become intertwined: (a) what goods and in what proportions is the charterer obliged to ship (b) how is the freight to be calculated (c) how are the damages to be calculated?   The first case was Thomas v Clarke and Todd (1818) 2 Stark 450, 171 ER 702, a decision at nisi prius.  The charterer had the option to ship various types of merchandise, for each of which a separate rate of freight was fixed.  In answer to a claim for a complete failure to ship, the charterer argued that the damages should be assessed on the basis of the cargo which yielded the least amount of freight.  Lord Abbott CJ did not accept this, and directed the jury that 'the proper course would be to estimate the freight by means of an average, so as to take neither the greatest possible freight nor the least'.   The second case is Capper v Forster (1837) 3 Bing NC 938, 132 ER 672.  The charter here was different from the one considered in Thomas v Clarke and Todd since it stipulated for the loading of 'a full and complete cargo of lawful merchandise', and then went on to enumerate freights for various types of goods.  The ship brought home a part cargo of goods which were not of the type described.  Holding that the damages should be assessed on the basis of an average of the stipulated rates, Tindal CJ cited Thomas v Clarke and Todd, and said (3 Bing NC 938 at 949--950, 132 ER 672 at 677): '. . . the original intention and expectation of the parties at the time the charter was entered into as to the amount of freight which would become payable for the voyage, must have been founded upon the assumption, that the ship would bring home a cargo consisting of all or some of the enumerated articles in both cases, therefore, there exists the same necessity of applying the rule above laid down, which necessity is grounded on the consideration, that unless you adopt this rule, you have no other guide whatever the liberty, ''to fill up with other lawful merchandises,'' being understood by us to mean other lawful merchandises, ejusdem generis, at least so far as the calculation of the freight is involved in that construction.' Finally, there was Cockburn v Alexander (1848) 6 CB 791, 136 ER 1459, where the charter stipulated for a full and complete cargo of wool, tallow, bark or other lawful merchandise 'with separate rates of freight for wool, pressed and unpressed, tallow, bark and hides with stated maximum quantities of bark, tallow and hides'.  The ship in fact loaded cargo which included quantities of tallow and bark greater than the stipulated maximum, and also certain other goods for which no rate of trade was fixed.  The question was whether, as the defendants asserted, freight should be paid in an amount which would have resulted from applying the market rate to all the unenumerated articles, whereas the plaintiffs said that the assumption should be made that the ship had carried the maximum of bark, tallow and hides, and had filled up the remainder with wool.  The Court of Common Pleas upheld the plaintiffs' argument.   The first two cases are not easy to explain, but I believe the reasoning to be as follows.  The courts concerned themselves not with the charterer's obligation to ship, but with his obligation to pay freight.  They held that whatever the charterer might choose to ship, and in whatever proportions, he was obliged, on the true construction of the charterparty, to pay freight calculated on the basis of an average of the enumerated articles.  Cockburn v Alexander was, however, to a different effect.  Here, the court did not assume any kind of average.  Instead the reasoning proceeded by two stages.  First, the freight clause was understood as requiring the calculation to assume that the unenumerated articles were ejusdem generis to those for which freight was fixed, and that the whole scheme for the calculation of freight must be found in the contract itself, without recourse to rates derived from the market.  Second, it was held that the freight for the notional full cargo composed of enumerated goods and nothing else should, for the purpose of assessing damages, be computed on the basis of a cargo distribution which would yield the lowest recovery to the plaintiff.  It was in relation to the first stage of the reasoning that recourse was had to Capper v Forster, as pointing out a solution to the problem of 'lawful merchandise' not expressly catered for in the freight clause.  The second stage of the argument was quite different, and indeed it seems to have been uncontroversial.  At this stage, the court departed from the precedent of the two earlier cases.  The question of striking an average does not appear even to have been considered.  Instead an approach was adopted which was entirely in line with modern doctrines.   In these circumstances, it seems to me plain that Cockburn v Alexander provides no authority for the plaintiffs' argument in the present case.  There remain the two earlier decisions.  I find it hard to be sure how these are to be understood.  Perhaps they are cases on the computation of freight.  If so, the subject matter of the contract was far too distant from the present to allow any reliable analogy to be drawn.  Perhaps again they were decisions which show that in that particular context the ordinary process of proceeding to construe the obligation before assessing the damages could be short-circuited by the application of a conventional measure of damage.  If so, I venture to doubt, notwithstanding the high authority of the judges concerned, whether even in its own limited sphere the convention can now be regarded as good law.  Even if it could, I see no reason to transfer it afresh into the present very different context.   It appears to me, therefore, that the choice must be made between two alternatives.  Either the contract is read precisely as written, requiring the defendants to buy at least 16,000 dresses each season, with the whole range of dresses to choose from without restriction, in which case the application of the ordinary rules demands that the cheapest dresses must form the yardstick for the entire computation.  Or it must be subject to an implied term that the choice must be made in a manner which is reasonable in all the circumstances.  If so, Abrahams v Herbert Reiach Ltd [1922] 1 KB 477 shows that a selection must be made from those methods of performance which can be regarded as reasonable, on whatever basis yields the result least unfavourable to the defendants.   Although the choice between these alternatives lies at the heart of the present dispute, it is not one which admits of lengthy discussion.  The plaintiffs say that, whatever its formal structure, the contract set up an arrangement in the nature of a joint venture, requiring the parties to co-operate in the effective exploitation of the plaintiffs' garments within the territory, an exploitation which would involve the protection of the plaintiffs' goodwill.  This objective would not be promoted by the sale of any random collection of dresses, far less a collection which by no stretch of the imagination could possibly have been sold, and which (it might well be) could not even have been manufactured in time.  On the defendants' side, cl 7 is relied on to show that they were given such a freedom of action that the plaintiffs must be taken to have ceded the control and protection of their goodwill, for so long as the contract was in force.  Again, the defendants say that the minimum obligation would, ex hypothesi, come into play at a time when the defendants would be unable to sell the purchased goods and could therefore not be expected to buy at any but the lowest possible prices.   I do not think that either of these arguments is an answer to the plaintiffs' claim.  Certainly, cl 7 did give a wide discretion as to volume, outlets and retail price, and the taking of a wrong view on these matters could cost both parties dear, and damage the goodwill for the future.  Nevertheless, I find it hard to accept as being in the contemplation of the parties, when the agreement was made, that the defendants could permissibly order 16,000 garments of the same size, style and colour, and no others at all, a configuration which would alienate their wholesalers, deprive the plaintiffs of anything but a ludicrous presence in the territory, and kill the market not only for the current seasons but for those which were to follow.  Even taking into account the very stringent tests imposed under English law for the implication of a term, some constraint on the defendants' freedom of choice must be assumed, in order to make sense of the agreement.   Nor in my judgment is it correct to answer that an implication cannot be made in face of the fact that the defendants would only be in breach of the minimum obligation clause if sales within the territory could not be made in the quantities which the clause required.  In my judgment, this approaches the problem from the wrong end.  The implication of terms must be assessed on the assumption of performance, not breach.  Performance consists of the purchase of 16,000 dresses, or more, if the defendants wanted them.  It was explicitly provided that the purchases were to be 'for the territory', not that the dresses were to be left in a warehouse, or put on a bonfire.  A purchase of 16,000 garments, and no more, would not be a measure of distress, but would be a full performance of the contract, and there is in my view no justification for distorting what would otherwise be a businesslike interpretation of the agreement, by assuming that the defendants would want, and should be allowed, to carry out that performance in a way which would do nothing but harm to the joint interests of the parties, and which would serve only the self-contradictory purpose of minimising damages which in a case of full performance would never fall due.   Accordingly I consider that the agreement must be construed as subject to an implied term that the garments would be selected in a reasonable manner.  Since selection would be a matter of judgment, this leaves open the very strong possibility that there would not be a unique reasonable selection, but a range of such selections, some yielding a greater total price than others for the 16,000 chosen garments.  On this basis, the damages should, as I have already suggested, be assessed in terms of that reasonable selection which would yield the lowest price.  I am very conscious of the problem which will face the court in establishing the boundaries of the range, since the sales in mitigation go some way towards showing what a reasonable selection would have been, but they do not show what other selections might also have been made which would qualify as reasonable.  I can only say that a similar, although less complex, task is one which the Court of Appeal found itself able to perform in Abraham v Herbert Reiach Ltd, on the basis of evidence much less comprehensive that is likely to be available at the hearing of the references to damages.   I now turn to a separate issue which relates to the manner in which the sales in mitigation should be brought into account.  I can deal with this quite briefly.   It was argued that the sales made by the plaintiffs directly should be treated as performance, pro tanto, by the defendants, and that only the balance of the 16,000 dresses should be the subject of a notional distribution and calculation of damages.  I do not accept this.  Sales by the plaintiffs were not in the nature of a substituted performance, for there was nothing left to perform.  There was a completed anticipatory breach of the whole obligation, followed by partial mitigation.  As in every case of damages, the correct approach is to compare the position which the plaintiff would have occupied if the contract had been fully performed, and the position which he actually occupied in face of the breach.  For the first half of the computation it is necessary to look at the whole of the 16,000 garments to see how the sizes, and so on, would have been distributed, and to work out the total price which would have been payable.  The second half of the computation must be made of those garments which were sold, or (if failure properly to mitigate is in issue) could have been sold in the territories reopened to the plaintiffs by the termination of the contract.   It seems to me that the two seasons for this purpose should be treated as a whole.  True, the court may have to look separately at the seasons when deciding what would have been a reasonable choice, but there was a single anticipatory breach in relation to the entire quantity of 32,000 dresses.  Both the breach and the subsequent mitigation must be looked at as a whole.  I can see nothing inconsistent here with the decision or the reasoning of British Westinghouse Electric and Manufacturing Co Ltd v Underground Electric Rlys Co of London Ltd [1912] AC 673, [1911--13] All ER Rep 63.   I hope in the judgment just delivered that I have given sufficient guidance to enable the parties now to proceed to the concluding stage of the dispute.  I cannot part with the matter without expressing the earnest wish that these respectable parties should now, having got so far, find it possible in the light of the guidance provided by the court to compromise their dispute rather than fight it to a conclusion. DISPOSITION: Judgment for the plaintiffs for assessment of damages by a master. SOLICITORS: Martin Clore & Co (for the plaintiffs); Michael Goldstone & Co, Woodford (for the defendants).