Metal Roofing and Cladding P/L v Amcor Trading P/L [1999] QCA 472

[1] McPHERSON JA: Polyvinyl chloride, or PVC, pipes are manufactured and widely used in Australia for the purpose of reticulating or transmitting water and other liquids. Such pipes are commonly made from PVC resin by suspension polymerisation, which is a chemical process that involves dispersing the resin in water, heating it with additives that combine with the PVC homopolymer in the resin, and then passing it through an extruder to make the pipes. The resin is a powdery substance which is supplied to the manufacturer in containers or bags. It is a derivative of petroleum and a by-product of that industry, which is produced locally in Australia; but quantities of PVC resin are also imported from Europe and Asia.

[2] Amcor, which is the plaintiff in the action and the respondent to this appeal, is an entity that sells PVC resin which it imports from overseas. Pipemakers, which is a division of the appellant defendant in this action, is a manufacturer and seller of PVC pipes. Late in 1993, the defendant began buying PVC resin from the plaintiff, which was then ?relatively new? to the business of supplying PVC resin in the Australian market. At first, the resin supplied by the plaintiff was of Korean manufacture; but, when supplies from Korea ceased to be available, the defendant began supplying PVC resin manufactured in Shanghai in China according to technology developed in Japan and referred to as the Shenitsu process. This product, which, when the plaintiff introduced it, was new to Australia in the sense that (according to the plaintiff?s sales representative) no one had used it here before, carried the code S1000, and, later on, WS1000; but these designations were internal forms of identification used by the Chinese manufacturer for its own purposes. They have no relevance to the issues here except that the appeal is concerned only with PVC resin bearing the designation S1000.

[3] In about May 1995, the plaintiff agreed to sell, and the defendant to buy, a quantity or quantities of PVC resin of Chinese manufacture for delivery in late June or early July 1995. The relevant purchase notes, sales orders and invoice, or a typical example in each instance refer to so many tonnes of "PVC resin S1000", to a price per kilogram, and to a total price for the particular consignment in question. The resin was delivered to the defendant in bags in Brisbane, and was applied by it in manufacturing pipes using the suspension polymerisation method.

[4] Sale by description. There is no doubt that this transaction, and others like it, amounted to a contract for the sale of goods "by description" within the meaning of s 16 of the Sale of Goods Act 1896. It was a contract to sell goods that were not "specific" but "unascertained" at the time of agreement, and it was one into which the description of the goods as "PVC resin" plainly entered. Those were the words used to identify the kind of goods to be supplied: Ashington Piggeries Ltd v Christopher Hill Ltd [1972] AC 441, 502 and Speedway Safety Products Pty Ltd v Hazell & Moore Industries Pty Ltd [1982] 1 NSWLR 255, 258. In the court below, the learned trial judge seems to have accepted this view of the contract; but, in any event, there can, in my opinion, be no doubt about its correctness. It follows that, under s 16 of the Act, there was an implied term that the resin sold should correspond with the description "PVC resin". It was those goods which it was under s 29 of the Act the duty of the plaintiff to deliver to the defendant.

[5] The questions on appeal are: (1) whether the resin designated S1000 delivered by the plaintiff to the defendant answered the description "PVC resin"; if not, (2) whether the plaintiff was in breach of contract, and liable to the defendant in damages; and, (3) in what amount. It is convenient to approach these questions separately although they do at several points tend to merge or overlap.

[6] PVC resin. The trial was conducted as a commercial cause, in which bundles of agreed documents were put in evidence, together with written statements from witnesses, of whom some, but not all, were cross-examined at the hearing. As to the question whether the goods delivered by the plaintiff consisted of PVC resin, several witnesses gave statements or oral evidence, or both. There is a measure of agreement among them that there is no such thing as 100% "pure" PVC resin. Contaminates or additives, including water, are invariably introduced in the course of manufacture, and some of their residue always remains in the PVC resin. Statements about "pure" PVC resin take account of this residue. Such residues are, however, not expected to be present in the residue in substantial percentages. In cross-examination, Mr Aylett, a consulting specialist in plastics who prepared a report and gave evidence for the plaintiffs, considered the question he was asked to address was, When does it cease to be PVC? He answered by saying that, for PVC homopolymer, he "certainly wouldn't expect less than 99.7 per cent" PVC. The introduction, for example, of excessive levels of vinyl acetate results in hydrolysis, which produces a co-polymer polyvinyl alcohol PVA, which evidently has a different molecular structure from PVC homopolymer. Vinyl acetate polymers have, according to Mr Scully, ?dramatically lower softening and melting points than PVC homopolymers?. Mr Scully, who, like Mr Aylett, prepared a report for the plaintiff, said in it that:

"6. Most of the additives used to make PVC resin (excluding most of the water), or their degradation products remain in the PVC resin. For suspension polymers, this means that commercially pure PVC resins contain at least 0.2% [to] 0.3% of 'non-PVC'. Included in this is up to 0.1% residual water. Some Japanese technology requires the use of much higher levels of additives to make PVC resin."

In cross-examination, he agreed that he regarded PVC, or "pure" PVC, as containing between 0.2 and 0.3 per cent non-PVC material; but said that he would "still call it" PVC if it contained a contaminant of as much as 3.2 per cent.

[7] Among others, Dr Rowan Truss, Associate Professor in the Department of Mining, Minerals and Materials Engineering at the University of Queensland, gave a written statement, on which he was not cross-examined, saying:

"3. I have no doubt that, from an academic or commercial view, the term 'PVC' means what it says. PVC may contain a small amount of chemicals which arise from the production of the resin, particularly from PVA which may be added to control particle size during production. However, I would expect PVC purchased in Australia to mean a product which is at least 99.9% pure PVC."

Dr Truss has many years of experience in the processing, structure, service and performance of polymeric materials. From 1982 to 1988, he was a research scientist, and later Principal Research Scientist, for the ICI Research Group, and has spent many years researching polymer structure, properties and failure. In oral discussions, of which a record became evidence at the trial, Mr Steve Tan, of European Vinyls Corporation (EVC) in Britain, who in 1996 had analysed a sample of the S1000 resin supplied to the defendant, was asked about the purity of PVC resin. He said he would expect to see impurities consisting of 0.3% volatiles, plus 0.1% for granulating agent and other necessary materials; or "about" 0.5% non-PVC. In a written report (ex 27) provided to the defendant and admitted in evidence at the trial, Mr LeHunt of ICI said that, in the PVC pipe industry, "the conventional assumption is that PVC resins received in processing plants consist of 100% PVC, and only PVC, as homopolymer"; but that, in practice, there would be some residual traces of catalyst material, which would be expected to be at a level less than 0.1% of the total content. He was not cross-examined on his statement to that effect.

[8] From this it appears that there is a range of opinions in scientific circles about the precise percentage of residual contaminants or additives expected to be present in PVC resin. Some will always be there; but, with the possible exception of Mr Scully, the expectation is that PVC resin will, according to Mr Tan, be at least 99.5% "pure"; or, according to the others, 99.7% "pure". Even Mr Scully in his report spoke of "commercially pure" PVC resins as containing only 0.2% to 0.3% of "non-PVC".

[9] A question is raised about what was meant by these witnesses when they spoke of an "expectation" concerning the percentage of contaminants in PVC resin. To my mind, it is obvious that they were referring to what a commercial user, and therefore a buyer, of PVC resin purchased in Australia would expect the composition of the PVC resin to be. This is plainly what was intended by Dr Truss, who in his statement (para 3) speaks of PVC resin purchased in Australia. Both he and Mr Harry Listoken refer to "off spec" resin, which Dr Truss described as resin "which does not comply or fall within the normal specifications of the manufacturer [but?] is marketed as such". In the experience of Dr Truss, "Australian manufacturers would always point out the fact that the resin was off-spec ...". To Mr Listoken, who is the managing director of Perex Pty Ltd, which imports and sells PVC resin, and who has over 15 years of experience in the trade, "off-spec resin" usually means that "it has a contaminant in it, something which shouldn't be there". He would classify a PVC resin containing an amount of polyvinyl alcohol as "basically contaminated off specification".

[10] There is an American standard for PVC resin, but an Australian standard for resin, as distinct from PVC pipes made from it, has not yet been developed or adopted. When, therefore, witnesses talk of the expectations of users or buyers with respect to PVC resins, and equally when they refer to "off-spec" resins, they must, I consider, be speaking in terms of local commercial or market usage. The learned trial judge said, that in his view, the evidence was "far from founding a custom binding on the plaintiff"; but if by this he meant a custom of the trade that was notorious, uniform and certain, as in Nelson v Dahl (1879) 12 Ch D 568, 575, from which a term is to be implied in the contract, then, with respect, I consider his Honour misconceived the function or significance of that evidence. What was being put forward at the trial was not a trade custom in that sense, but simply evidence of the use in the Australian market of the words or abbreviation PVC resin to describe the kind of polyvinyl chloride resin that was being bought and sold as such in that market. To identify a subject matter being bought and sold, a description of some kind must be used and, except where it refers to a popularly known substance which needs no elucidation, evidence of what is commonly meant by that description in the trade is admissible to show what is meant by those who use it.

[11] In Nichol v Godts [1854] EngR 588; (1854) 10 Ex 190; 156 ER 410, where evidence was admitted to explain what was meant in the market by "foreign refined rape oil", the plaintiff seller contended that what had been delivered was known in the market as such, even though it in fact contained an admixture of hemp oil. The Exchequer Report of the case [1854] EngR 588; (10 Ex 191, 192) records that, on this point at the trial, "evidence was adduced on both sides". In the Law Journal Report, Parke B is quoted as having, in argument [1854] EngR 588; (23 LJ Ex 314, 315), said that at the trial (at which he presided):

"Evidence was given on each side as to this oil being refined rape oil, and that question I left to the jury."

In Nichol v Godts, Parker B told the jury that if the plaintiff could show that, by the "general usage of merchants", the oil in question was denominated foreign refined rape oil, the plaintiff seller was entitled to a verdict for the price. The jury having found a verdict for the defendant buyer, an order nisi for a new trial was discharged by the Court. Benjamin's Sale of Goods s11.004 (5th ed) mentions Nichol v Godts as one of the decisions on which the codification in s 13 of the Sale of Goods Act (UK) (s 16 of the Queensland Act) was based. See also Josling v Kingsford [1863] EngR 213; (1863) 13 CB (NS) 447; 143 ER 177 (contaminated oxalic acid), where Erle CJ told the jury, correctly as it was held, that the question was whether what was delivered might in "commercial language" be said to come under the denomination "oxalic acid" (13 CF (NS) 447, 451; [1863] EngR 213; 143 ER 177, 179). A much more recent instance, on which the defendants also relied on this appeal, is Marimpex Mineralöl v Louis Dreyfus et Cie Mineralöl [1995] 1 Lloyd's Rep 167, where Clarke J admitted evidence of what was accepted in the trade as the limit of bacterial contamination in a substance sold as "normal Russian gas oil", a description which his Lordship found was not satisfied by oil contaminated beyond the level accepted in the trade as satisfying that description.

[12] In my opinion, on the evidence of the witnesses in the present case, the defendant succeeded in proving that in the Australian market at the relevant time the description "PVC resin" meant resin that was at least 99.7% pure PVC and to that extent at least free from contaminants or additives.

[13] Breach of condition. The next question is the extent to which, if at all, the defendant established a breach of the condition under s 16 of the Act that the goods must correspond with the description PVC resin. Various tests were carried out by different experts on the resin designated S1000, as well as that designated WS1000. According to one test conducted by Chemical Laboratories in Western Australia, a sample of the S1000 resin showed a contaminant or ?non PVC? content of as much as 3.2%; but Mr Aylett rejected the method of analysis used there as unreliable. His Honour said there was "room for differences" in determining the polyvinyl chloride content of resin, "which is an exacting task requiring the use of sophisticated equipment". He said the evidence was "also open to the conclusion" that the chlorine microanalysis process used in some of the tests is "a more accurate method, and it establishes that the resin in question was 99.5% polyvinyl chloride". His Honour accordingly found that the defendant had not proved the breach of contract on which it relied in the action.

[14] In finding that the chlorine microanalysis process showed that the resin delivered by the plaintiff was 99.5% polyvinyl chloride, the learned trial judge appears to have confused the results obtained by that process of testing resin designated WS1000 with results of testing resin designated S1000. It was resin WS1000, and not resin S1000, that produced a result of 99.5%. For the S1000 resin, the chlorine microanalysis carried out by or at the behest of ICI (now Orica), as disclosed in the Probe Analytical Report dated 1 April 1998, records the results for total PVC reported by Dr Ramon Tozer as being: A) 99.30%; B) 99.21%; and C) 98.53%, giving an average for the three of 99.01%. Dr Tozer also reported, however, that sample C (98.53%) was "noted to have considerably more cloudiness in the mother liquor than samples A) and B)", and that, as a result, "it would be reasonable to assume this value [to be] erroneous".

[15] It appears that his Honour was prepared to accept the chlorine microanalysis testing method and results, and that, if it had not been for the confusion with the WS1000 result, he would have found that the average result for S1000 was only 99.01%. I consider that a conclusion to that effect should now be accepted on this appeal. If result C) is excluded as erroneous, the average result derived from tests A) and B) in Dr Tozer's report would still be resin that was less than 99.26% pure PVC. Mr Aylett, who, it will be recalled, was the plaintiff's principal expert witness at the trial, said that he would not expect less than 99.7% for PVC resin homopolymer. On this footing, the defendant established that the resin S1000 delivered by the plaintiff failed to correspond with the description "PVC resin", and so was in breach of a condition of the contract.

[16] In cross-examination, Mr Aylett sought to counter this conclusion by saying that, although "nominally" the expectation would be "say, 99.6 or 99.7%, every PVC producer has the right to develop a competitive product in the market". This may, of course, be accepted. The question is, however, not what a PVC resin manufacturer is entitled to produce or develop, but what a buyer in the Australian market would expect to receive on buying resin described as PVC resin. While accepting that "nominally" the expectation would be 99.6 or 99.7%, Mr Aylett said he would "not assume anything" but "would need to process and check that material". If, by that, he meant that a buyer in the Australian market at the relevant time of something described as PVC resin was obliged or expected to test for contaminants before using or buying that substance, then, with respect, it is not an approach that is or ought to be sanctioned in this branch of the law. Testing by chlorine microanalysis appears to be a time consuming process and is no doubt expensive. Buyers whose business is manufacturing and selling pipes cannot be expected to engage in that kind of investigation preparatory to buying or using what is described as "PVC resin" in a market where that description has an accepted and widely understood meaning. If an importer like the plaintiff Amcor introduces into the Australian market a new product that does not conform with the accepted description or meaning, it is incumbent on it, when selling that product, to disclose those particulars in respect of which it does not so conform. See Pinnock Brothers v Lewis & Peat Limited [1923] KB 690, 698. There is, as it happens, nothing to suggest that the plaintiff carried out tests of the kind suggested by Mr Aylett before selling S1000 in the Australian market as PVC resin.

[17] Measure of Damages. This leaves for consideration the question of damages, which in some ways is the most difficult of the issues on the appeal. In the action against the defendant, the plaintiff obtained judgment for a total of $781,237.89, of which $589,086.50 was for the price of goods sold and delivered. The balance is represented by interest on that sum and damages for non-acceptance, of which more will be said later in these reasons. The judgment amount has since been paid by the defendant. The principal issue on the appeal now is the defendant's counterclaim for damages for breach of the condition implied by s 16 of the Act, which is what has been considered so far. At the trial, there was a counterclaim for damages based on the size of the bags in which the plaintiff supplied resin to the defendant; but his Honour rejected it, and it has not been pursued on the appeal.

[18] Turning to the question of damages for breach of the condition, s 14(3) compels the defendant in circumstances like these to treat it as a breach of warranty and not as a ground for rejecting the goods. Section 54(1)(b) then enables the buyer to maintain an action against the seller for damages. See Wallis, Son & Wells v Pratt & Haynes [1911] AC 394, 395. Section 54(3) provides that, in the case of breach of warranty, the measure of the loss is "prima facie the difference between the value of the goods at the time of delivery to the buyer, and the value which they would have had if they had answered the warranty". Although s 54(3) refers specifically to breach of warranty "of quality", the expression "quality of goods" is defined in the Act to include their state or condition. The fourth edition of _Benjamin's Sale of Goods s_1305, 1307 note 51, considered that s 54(3), or its English equivalent s 53(3), provides the measure of damages for breach relating to the description of goods. Subsequent legislation in England may explain why more recent editions of Benjamin have to an extent retreated from that position; but certainly in Wallis Son & Wells v Pratt & Haynes [1911] AC 394, their Lordships showed a general disposition to apply the provisions of s 53 (s 54 of the Queensland Act) to a breach of the condition under s 13 (Qld s 16); and in Marimpex Mineralöl v Louis Dreyfus et Cie [1995] 1 Lloyd's Rep 167, 179 col 2, the damages were calculated by applying the prima facie measure specified in s 53(3) (Qld s 54(3)). See also Bostock & Co v Nicholson & Sons Ltd [1904] 1 KB 725.

[19] The specific reference to quality in s 54(3) of the Act may, in the end, not count for much because s 54(3) is no more than a particular application of the more general principle in s 54(2), which is that the measure of damages for breach of warranty is the estimated loss directly and naturally resulting, in the ordinary course of events, from that breach. This, it is universally accepted, is simply a statement of the general measure of damages for breach of contract laid down in Hadley v Baxendale and refined in later decisions. See Bostock & Co v Nicholson & Sons Ltd [1904] 1 KB 725, 735. Under Roman law and systems that follow it, a buyer's primary remedy for breach is to return the goods and get his money back. Relief in that form was quite unsuited to the world-wide system of trade in which our law of sale of goods evolved in 19th century England. English law, and consequently our own, has tended to limit the right to reject, as for example in s 14(3) of the Act, and to confine the buyer's remedy to monetary compensation in the form of damages for breach of warranty measured by reference to the prevailing market price. It is only in this way that a buyer like the defendant in this case can be compensated for having paid more than the market ?value? of goods which, because of their non-conformity with description, would otherwise have attracted a substantially lower market price. Damages recoverable by a buyer who is entitled to, but does not, reject goods are measured by reference to the "value", or price at which those goods could, with reasonable diligence, have been sold on the market. See s 52(3); Kwei Tek Chao v British Traders and Shippers Ltd [1954] 2 QB 459, 497-498; and Bostock & Co v Nicholson & Sons Ltd [1904] 1 KB 725, as well as Loder v Kekule [1857] EngR 734; ; , which is an early decision holding that the measure in circumstances like these is the difference between the contract price and their value as so ascertained.

[20] For the plaintiff on appeal, Mr Savage submitted that to compensate the defendant buyer in that way ignored the circumstance that the defendant had in fact used the S1000 resin, or most of it, in the manufacture of pipes which it sold. As it happens, there is some material suggesting that the defendant may yet become a target of claims from persons who purchased from it pipes manufactured from S1000 resin. Whether that will in fact eventuate because of contaminant or PVA in the pipes depends on how much, if at all, or how fast, those pipes decompose or deteriorate in future. Claims by buyers of pipes manufactured by the defendant may conceivably be made only after the period of limitation for a claim by the defendant against the plaintiff has already expired. In any event, what the defendant in fact did with the resin is not critical to the defendant's claim against the plaintiff. It might, for a variety of reasons, such as financial problems, or overstocking, or a dearth of orders for pipes, equally well have chosen or been compelled to sell the S1000 resin, or some of it, on the open market. Its value for present purposes would then have been ascertained at the prevailing market price whatever it may have been. Mr Savage's submission really confuses the obligation of the seller under s 16 of the Act (correspondence with description) with its obligations under s 17(1) relating to fitness for purpose of the goods, and s 17(2) relating to merchantable quality. It is, in a legal sense, the same mistake Mr Scully was making when he said or implied that use of the Japanese technology resulting in a higher PVA content might nevertheless enable pipes to be produced that were fit for use as such. The legal obligations implied against a seller as conditions by s 17 are different, and serve different functions in the sale of goods, from the obligation implied by s 16. Even if the S1000 resin was merchantable, or reasonably fit for its purpose (which perhaps it may not have been), it nevertheless failed to correspond with the description under which it was sold. The point is covered by the decision in Arcos Ltd v E A Ronaasen & Son [1933] UKHL 1; [1933] AC 470. The dimensions of the goods sold there (timber staves for making barrels) did not correspond with the contract description. That the goods were merchantable under the contract was, according to Lord Buckmaster ([1933] AC 470, 474), "no test proper to be applied in determining whether the goods satisfied the contract description". So also Lord Atkin, who, in speaking of the finding that the goods were in fact fit for the particular purpose required, said ([1933] AC 470, 480):

"The implied condition under s 14, sub s 1, unless of course the contract provides otherwise, is additional to the condition under s13. A man may require goods for a particular purpose and make it known to the seller so as to secure the implied condition of fitness for that purpose: but there is no reason why he should not abandon that purpose if he pleases, and apply the goods to any purpose for which the description makes them suitable. If they do not correspond with the description there seems no business or legal reason why he should not reject them if he finds it convenient to do so."

According to the decision in Bostock & Co v Nicholson & Sons Ltd [1904] 1 KB 25, 734-735, the damages are different in such a case.

[21] Proof of loss. The question, then, is, whether the defendant proved damages within the meaning of the rule as stated. It must be said that there is not a great deal of evidence on the point. In a way, however, that is not altogether surprising. It may be expected that manufacturers of PVC pipes do not ordinarily wish to buy resin that is less than commercially pure as containing less than 99.7% PVC. As has been said more than once in these reasons, the plaintiff's Chinese manufactured product was new to the market.

[22] Given that the measure of loss for breach of the condition is the difference between the contract price and market value of the non-conforming goods, the next matter for consideration is whether the defendant established what that value was. The evidence on which the defendant relied for this purpose is limited to what was said by Mr Listoken in his statement ex 46, para 8, and in his testimony at the trial. Paragraph 8 of ex 46 is as follows:

"8. If PVC resin did contain a contaminant greater than 1%, as in this case between 2% and 3%, then in my experience it would be discounted in the market place by about 20 to 25%."

His evidence in chief at the trial, so far as relevant to this point, was:

"If one can use perhaps more concrete examples, if PVC resin contains three per cent polyvinyl alcohol, are you able to say at what discount it would sell in the market place? --- I would estimate somewhere between 30 and 50 per cent."

He was reminded that in his original statement (ex 46) he had estimated the discount for a contaminated product as 20 to 25% but was now saying 30 to 50%. His response was that his original estimate was based on a figure "off the top of my head"; but he had since examined the paperwork for some of his company's off specification materials and found it was sold at up to 50% discount.

[23] On appeal Mr Savage pointed out that in cross-examination Mr Listoken had said that the most common form of contaminant in PVC resin consisted of black carbon particles resulting from overheating in the resin manufacturing process. His evidence about the extent of the market discount was not, however, confined to that particular form of contaminant. What is a difficulty for the defendant is that Mr Listoken spoke of such a price reduction or "discount" only in relation to contaminants amounting to 1% or more. In para 8 of ex 46, he referred to PVC resin containing contaminant "greater than 1%". In evidence in chief he was asked about PVC resin containing 3% polyvinyl alcohol. Immediately after that he was asked a further question about the market discount on resin with polyvinyl alcohol this time reduced to 1%. As to that, he said:

"Well, I would estimate it would be the same because, if something is contaminated, it's contaminated, and you don't know until you begin to use it whether there is a problem with it or not. Therefore to get somebody to use it you have to give them a financial incentive to do so."

The problem remains that what Mr Listoken said about market discount or reduced price is in each instance referable to PVC with contaminant amounting to 1% or more. At best for the defendant, the Probe Analytical chlorine microanalysis tests arranged by Dr Tozer established a PVC purity level averaging 99.01%; but this average, it will be recalled, was arrived at by including test C) of 98.53%, which Dr Tozer thought it reasonable to assume was erroneous. If that result is disregarded, the average purity of the resin, as shown by the other two tests, is increased to 99.26%. Thus Mr Listoken's evidence about the reduction in the market price payable for contaminated resin does not quite match up with the percentage contaminant proved to have been present in the PVC resin S1000 sold by the plaintiff. He was saying the market price was reduced on account of the presence of contaminant amounting to or greater than 1% of the resin. The best average test result for the defendant is 99.01% or, perhaps more reliably, 99.26% pure PVC, or slightly less than 1% contaminant in the resin.

[24] To some, it might not seem much of a difference; but, in assessing damages for an alleged departure from, or non-correspondence with, a contract description such as PVC resin, it may not be possible to overlook even as fine a margin as that. It is true that, as Mr Listoken's evidence demonstrates, it is the presence of contaminants that causes prospective buyers to become wary; but there is no direct evidence of any impact on market price unless and until the percentage of contaminant present in the resin reaches a level of 1% or more.

[25] That, however, is not quite the end of the defendant's case. The test conducted by Mr Tan of EVC in Britain showed the presence of polyvinyl alcohol "at about the 2% level" in the sample of PVC resin analysed by him in 1996. His report (ex 42), which was admitted without objection at the trial, goes on to say:

"It is difficult to explain why the PVC polymer should have PVA in it; whilst PVA is used as a granulating agent in the PVC polymerisation process, it is normally not used in such vast quantities."

In his section of what was a court-directed joint report with Mr LeHunt, Mr Aylett said he could not tell what test method had been used by EVC, or (having regard to the use of the phrase "about 2%") the precise level of PVA polyvinyl alcohol in the sample tested. However, when asked in cross-examination whether EVC would have got the testing right, Mr Aylett said, "Yes. I have stated that EVC and ICI are the only two reliable ones". He appeared to think it possible that the presence of such a large amount of PVA as 2% might have been the result of an accidental inclusion of that material; but added that a purchaser of "new resin" containing PVA would need to work out how to "formulate" with that resin. However, no PVC pipe maker in Australia of whom Mr Aylett is aware was adding polyvinyl alcohol to the extent of more than .05%, and then only for the purpose of carrying out ?trial? work with that form of product.

[26] Partly no doubt because of the confusion over the WS1000 test results, his Honour did not make any finding about the status of the EVC analysis or result. It is, however, difficult to see how he could properly have rejected that result in view of Mr Aylett's evidence at the trial that the EVA test was reliable. Mr LeHunt plainly held such a view of the EVA test result. Since by Rule 766(1)(a) of the Uniform Civil Procedure Rules, the Court of Appeal has all the powers and duties of the court that made the decision appealed from, I consider that, on the uncontradicted evidence, we should now find that the PVC resin supplied by the plaintiff to the defendant in 1995 contained a polyvinyl alcohol contaminant of about 2% as revealed by the EVC analysis conducted in May 1996. Why the Probe Analytical tests of 1 April 1998 should have produced an average result of only 1% contaminant is not explained in the material. The sample used for those tests came from a single bag of S1000 resin which the plaintiff found in its warehouse apparently only shortly before the trial; and the analysis became available only after the trial had commenced on 31 March 1998. If the explanation was, as Mr Aylett may have been suggesting, that the EVC tests were conducted on a quantity of resin which had been accidentally over-contaminated with PVA in the course of manufacture, it nevertheless remained a breach of the condition as to description. The problem with which the defendant would then have been faced would have been its inability, without having each bag analysed, to determine which of the bags did, and which did not, contain resin contaminated to the extent of 2% or more or less. It could scarcely be expected to test or analyse the contents of each and every bag delivered before using it to manufacture PVC pipes.

[27] The result in my opinion is that the defendant has established a breach of contract on the part of the plaintiff in supplying, under the description PVC resin, S1000 which contained polyvinyl alcohol to the extent of some 2%. Since Mr Listoken's uncontradicted evidence is that the market price for PVC resin with that level of contaminant would be reduced by 20% to 25%, or even by as much as 50%, compared with the market price for 99.7% "pure" PVC resin, it follows that the defendant succeeded in proving actual and not merely nominal damages for breach of contract. Its loss arose from having paid the plaintiff more than the market price for resin of the kind or description in fact delivered under the contract for sale of PVC resin. Mr Savage suggested there was no evidence of the figure from which 20% to 25% or more was to be deducted or reduced; but the purchase price of the resin contracted to be delivered by the plaintiff is evidence of the market price for PVC resin answering the contract description. Mr Keane QC provided a schedule calculating damages and interest based on a 25% reduction from the contract price of the non-conforming S1000 resin supplied. He submitted, and it may be accepted, that it was not unreasonable to adopt 25% rather than 20% having regard to Mr Listoken's later researches showing a reduction of 50% in market price for contaminated PVC. There is no evidence to the contrary. Indeed, there is some evidence of sale at a loss of resin repossessed by the plaintiff in December 1995, although its relevance (if any) to the defendant?s counterclaim was not investigated at the trial. Twenty five per cent of the total amount paid by the defendant for S1000 resin delivered by the plaintiff is $499,407.50, to which interest at 10% has been added, to produce a total, calculated to 23 September 1999 when the appeal was heard, of $220,993.26 = $720,400.76. The accuracy of the calculations are not contested. The total sum for which the defendant is entitled to judgment on its counterclaim is therefore $720,400.76, which represents the amount by which the defendant was (to use counsel?s term) ?overcharged? for the resin in fact delivered.

[28] The alleged compromise. There is a further issue requiring attention on the appeal. In the action the plaintiff also recovered judgment for an amount of $60,879, together with interest, as damages for the defendant's non-acceptance or rejection of a quantity of 308 tonnes of resin sold and delivered by the plaintiff in September 1995. On that occasion the resin in question was not S1000 but WS1000. On the evidence and the findings of the trial judge concerning the level of purity of WS1000, the defendant had no right to reject this consignment; but it nevertheless did so. A dispute followed in which the plaintiff refused to accept the return of that consignment of resin. It did so only after what the pleadings call a second agreement had been reached. One term of the agreement was that payment for the invoices in respect of it should be postponed for 90 days after the last day of the month in which the goods were delivered, which took it to a date in December 1995. Before that date had arrived, the plaintiff with the consent of the defendant retrieved the resin in question, or part of it, under an arrangement by which the plaintiff issued a credit note for the amount of the invoices involved. In December 1995 the plaintiff resold the repossessed resin, or 144.95 tonnes of it, at a loss of $60,879, which is the amount for which damages were awarded in favour of the plaintiff. It had and has a right to those damages for non-acceptance or rejection by the defendant of the resin delivered in September 1995 unless it agreed to compromise or forego its claim to them.

[29] The plaintiff claimed to be entitled to repossess and resell the resin pursuant to clauses K, L and Q of its General Terms and Conditions of Contract, which it was submitted were incorporated into the sales of resin to the defendant. The learned trial judge found that, by a course of dealing between the parties, those Conditions did become part of their contract. The defendant challenged this conclusion on appeal; and also submitted that, even if those conditions formed part of the contractual arrangement between the parties, the plaintiff had, in any event, expressly agreed to forego its right to rely on the rights conferred by those clauses.

[30] The defendant's submission proceeds as follows. There was a dispute about the plaintiff's right to repossess the September consignment of resin. Having agreed to postpone payment for 90 days, the time for exercising its right of repossession under clauses K, L and Q had not arisen and did not arise until December 1995. In the meantime, the plaintiff sent a letter dated 21 November 1995 to the defendant stating that, in consideration of the defendant's consenting to that repossession, the plaintiff would "issue a credit note for the invoiced value of all stock repossessed". The letter from the plaintiff asserted that its repossessing of the stock was being conducted "pursuant to its contractual rights" under clauses K and L of the General Conditions. On the same day the defendant, by a fax dated 21 November 1995 to the plaintiff, advised its consent to the plaintiff repossessing the resin in question "on the basis that a full credit for the original invoiced value of the resin would be provided by the plaintiff without any deduction". The second paragraph of that fax went on to say:

"Pipemakers will not accept any responsibility for losses incurred by Amcor in subsequently disposing of the faulty resin."

The resin was subsequently repossessed by the plaintiff on dates between 24 November and 4 December 1995, and resold at a loss, giving rise to the claim for damages for which judgment was given for the plaintiff in the sum of $60,879 and interest.

[31] What is submitted is that the plaintiff abandoned any claim it might have had to recover those damages when it repossessed the resin on terms which included the second paragraph of the defendant's fax dated 21 November disclaiming responsibility for losses incurred by the plaintiff on resale of the resin. The fundamental difficulty with that submission is that it does not appear whether the defendant's fax of 21 November followed the plaintiff's letter of that date, or preceded it. The sequence is critical. If the plaintiff's letter was despatched and received first, then the defendant's fax amounted to a counteroffer on terms which, it might be inferred, the plaintiff accepted by conduct when it proceeded to repossess the resin on those terms. The opposite result would ensue if the fax contained the offer and the plaintiff's letter a rejection of that offer itself constituting a counteroffer which was accepted by the defendant by allowing the plaintiff to repossess.

[32] It is not possible to resolve this question on the material on the record. The submission that the plaintiff compromised and surrendered its right to damages for non-acceptance or rejection of the resin was not raised in the pleadings on which the action was tried at first instance. In consequence, according to Mr Savage, the issue was not explored at the trial. The defendant should not now be permitted to raise it on appeal. To do so would require an amendment to the pleadings, which was not sought by the defendant. In accordance with Suttor v Gundowda Pty Ltd [1950] HCA 35; (1950) 81 CLR 418, 438, the plaintiff is entitled to object to its being raised for the first time on appeal "if evidence could by any possibility have prevented the plaintiff from succeeding". In a matter in which an inference of acceptance by conduct of the defendant's faxed counteroffer (if that is what it was) would need to be drawn, the possibility of evidence capable of displacing that inference cannot be discounted. Without having the opportunity of adducing such evidence, the plaintiff cannot be found to have agreed to compromise, surrender or forego its right against the defendant to recover damages for wrongful rejection of the September consignment of WS1000 resin. By s 29 of the Act, the defendant as buyer was under a duty to accept and pay for the goods. For a breach of that duty the plaintiff is by s 51(1) entitled to maintain an action for damages for non-acceptance, the amount of which falls to be ascertained under s 51(3) by the difference between the contract price and the market price. The defendant is therefore not entitled to succeed on this aspect of the appeal. The award of $60,079 as damages together with interest as part of the judgment in favour of the plaintiff must therefore stand.

[33] The exemption clause. The General Terms and Conditions put forward by the plaintiff contains the following clause D headed Warranties:

"(a) Any advice, recommendation, information, assistance or service provided by Amcor Trading in relation to the goods supplied by it or their use or application is given in good faith, as believed by Amcor Trading to be appropriate and reliable, however, it is provided with a disclaimer for any liability or responsibility on the part of Amcor Trading.

(b) The Customer accepts all risk and responsibility for consequences arising from the use of goods whether singly or in combination with other products.

(c) All warranties, guarantees and conditions, other than those expressly stated in the Contract, and whether implied by statute, common law, custom of the trade or otherwise are, to the extent that the law permits, expressly excluded".

On appeal the plaintiff did not, either in its written outlines or oral submissions, seek to rely on the provisions of clause D as exempting it from the obligation to deliver resin corresponding with the contract description. The defendant nevertheless considered it necessary to address the issue on appeal, and, because of that, it is desirable to consider the matter here.

[34] It is now well settled that in Australia the meaning and applicability or otherwise in particular circumstances of an exempting provision like that in cl D depends on its language, read in context ?and not on any a priori notion that non-delivery of goods was not intended to be protected?. See Nissho Iwai Australia Ltd v Malaysian International Shipping Corporation Berhad [1989] HCA 32; (1989) 167 CLR 219, 227, which was, however, it may be noted, a case of carriage by sea and not sale of goods. I think it is plain that the plaintiff could not rely on either paras (a) or (b) of cl D of the General Conditions. If it is protected from the defendant?s claim by any of the provisions of that clause, it can only be by para (c). The defendant?s response is that what it is complaining about here is a breach not of an implied condition, but of a condition ?expressly stated in the Contract?, namely the description ?PVC resin? under which the goods were sold to it. It must at once be said that s 16 does in terms treat the requirement of correspondence with description as an ?implied condition?. This nomenclature is, however, widely acknowledged to be a weakness in the drafting of the Act. See Benjamin?s Sale of Goods ??11-02 to 11-04 (5th ed).

[35] Failure to deliver the goods contracted to be sold or, what is the same thing, delivery of goods that do not answer, correspond or conform to the contract description, is a breach not of an implied, but of an express condition (if it is properly considered to be a condition at all) or term of the contract. See Pinnock Brothers v Lewis & Peat Ltd [1923] 1 KB 690, 698. In the present case, the form of the exemption conferred by para (c) of cl D is almost identical with, and probably modelled on, that in Andrews Brothers (Bournemouth) Limited v Singer & Co Ltd [1934] 1 KB 17, which, like this, expressly excluded ?all conditions ? whether imposed by statute, common law or otherwise?. In holding that such an exemption did not protect a seller who delivered a Singer car, which had been driven some 550 miles, instead of the ?new Singer car? required by the contract, Scrutton LJ said ([1934] 1 KB 17, 23):

"Where goods are expressly described in the contract and do not comply with that description, it is quite inaccurate to say there is an implied term. The term is expressed in the contract".

Greer LJ agreed, saying (at 25) that the phrase ?new Singer car? was an express and not an implied term. His Lordship explained the form of s 13 (Qld s 16) of the Act as follows:

"It may be right to say that the descriptive terms of the contract, though they are express terms, are not expressly made conditions of the contract, and by using the expression ?implied condition? in s 13, those responsible for the statute did not mean to say that the obligation to supply the thing described is not an express obligation; they merely meant to say that the description is not a mere term or a mere warranty but is a condition of the contract by implication of law".

Having, as the plaintiff did, in cl D (c) of the General Conditions used a verbal formula so close in its terminology to that in Andrews v Singer, the only reasonable conclusion that can be drawn from its having done so is that it was intended that the decision in that case should apply to it. The clause in question therefore did not have the effect of excluding the express obligation of the plaintiff as seller in this case to supply PVC resin to the defendant. Indeed, that obligation is one that is ?expressly stated in the contract?, to which the exempting provisions of cl D (c) are specifically declared not to apply.

[36] It should be noticed that the first paragraph of the General Conditions expressly recites that those General Conditions:

"? comprise the sale terms and conditions of the contract between Amcor Trading and the Customer (?the Contract?) notwithstanding anything that may be implied or expressed to the contrary on the Customer?s orders or inquiries ?".

This may represent an attempt by the drafter to achieve the same result as that reached in the later decision of L?Estrange v F Graucob Ltd [1934] 2 KB 394, where it was provided that the agreement contains ?all the terms and conditions under which I agree to purchase the machine specified above ??. If so, I consider that the attempt fails. The General Conditions in this instance speak simply of ?goods? or ?the goods? without anywhere condescending to describe what they are. To find out what they are, it is necessary to refer to documents and conversations passing between the plaintiff and the defendant in this case, and in particular the plaintiff?s sales orders and invoices, which speak of ?PVC resin?. Without referring to those conversations, sales orders or invoices, it would be impossible to identify ?the goods? with which the parties and the General Conditions are concerned. The standard General Conditions being used by the plaintiff are no doubt employed for a wide variety of different kinds of goods in which it trades, which would explain why the goods in question are not specified in the General Conditions as such; and, unless it be assumed that the parties had no goods and hence no contract in mind, looking outside the General Conditions is essential in order to identify the goods referred to. The expression ?terms and conditions? in the first paragraph must therefore refer to matters other than the description or specification of the goods agreed to be bought and sold. As it is, the expression ?General Terms and Conditions of Contract? itself implies that there are special terms applicable to the particular contract of sale of goods that is taking place. The result is that the General Conditions, and in particular cl D (c), did not exclude the obligation of the plaintiff to perform its contract to sell and deliver PVC resin. It may be added that L?Estrange v F Graucob Ltd seems to have assumed without deciding that the object delivered could, even though it worked only sporadically, properly be described as a "machine".

[37] Incorporation of General Conditions. The plaintiff?s General Conditions have so far been referred to as if they formed part of the relevant contracts or sales between the parties. In the court below, the learned trial judge found that, after the defendant?s first order for PVC resin was placed with the defendant in January 1994, ?a course of dealing? evolved under which, in response to the defendant?s orders, the plaintiff supplied further quantities of resin using documentation containing its standard printed General Conditions. On behalf of the defendant on appeal, it was submitted that a course of dealing which incorporated those General Conditions was never established. The defendant?s submission was based on the fact that the conditions in question were printed on the back of each invoice, which was sent to the defendant only after the relevant resin had been ordered and the contract with respect to it had been made.

[38] The plaintiff had, however, a practice of requiring a new customer to sign a written application form for credit addressed to it by that customer. Such an application form was signed by or on behalf of the defendant on 2 December 1983. It contains various terms relating to the provision of credit, but concludes with the following words:

"I/we acknowledge receipt of and accept the General Terms and Conditions of Sale ??.

The defendant relied on the fact that, contrary to that acknowledgment, the General Terms and Conditions of Sale were not received by it on or before that form of application was signed. For my part, I cannot see that it matters. The plaintiff is not here attempting to set up an estoppel against the defendant based on its having signed the credit application containing the acknowledgment of receipt of those General Conditions. What matters is that the defendant accepted the general conditions and signed the form of application as having agreed to do so.

[39] Despite the use of the word ?accept?, it is, I think, clear enough that technically the concluding words of the credit application embodied an offer to the plaintiff from the defendant to carry on their buying and selling on the terms of the General Terms and Conditions of Sale. That offer was certainly accepted at the latest when, if not before, the plaintiff first supplied resin at the request of the defendant. It is true that the defendant did not, or may not have, become aware of the terms of those General Conditions until the first invoice was delivered by the plaintiff, or even at all. But that is irrelevant. It is perfectly possible in law for a person to contract on terms that are incorporated into a contract only by reference to another document, which that party has never before seen, and may in fact never set eyes on. A well known if not very attractive illustration is provided by Thompson v London, Midland & Scottish Ry Co [1930] 1 KB 41. There the relevant incorporated terms were contained in railway time tables (which could be obtained only by payment on application to the defendant) that were referred to in printing on a ticket bought on the behalf of the plaintiff. Although she was blind, and consequently knew nothing about the printing on the ticket or the terms, she was nevertheless held to be bound by the provisions in them.

[40] Here the credit application was signed by the defendant, who in doing so agreed to be and became bound by those General Conditions once the application for credit was accepted by the plaintiff. All that remained for the plaintiff at the trial was to identify the General Conditions referred to in the credit application. On the evidence, the only such General Terms and Conditions of Sale to which the application for credit or agreement was capable of referring were those described by Mr Nelson at the trial, which appeared on the backs of the invoices sent by the plaintiff to the defendant. They were thus sufficiently identified as the Conditions being referred to in the application for credit that was signed by the defendant, and they were therefore effectively incorporated in the contracts or sales that subsequently took place between the parties : cf Smith v South Wales Switchgear Co Ltd [1977] UKHL 7; [1978] 1 WLR 165, 176-177. As it happens, apart from the matter of the exemption clause D (c), to which reference has already been made, the relevance of the General Conditions is limited to the provisions in cl K affecting interest and costs. On this issue, I would not disturb his Honour?s decision.

[41] The result is that in my opinion the defendant?s appeal against the judgment on the claim in the action should be dismissed. The appeal against the judgment dismissing the counterclaim should be allowed and that judgment set aside. On the counterclaim there should be judgment for the defendant for $720,400.76.

[42] The question of the costs of the action and counterclaim presents some difficulties. Most of the trial was taken up with proof of matters relating to the counterclaim on which the defendant has, according to these reasons, now succeeded. It may be that the plaintiff should recover the costs of its claim in the action, and the defendant its costs of and relating to the counterclaim: cf McDonnell & East Ltd v Macgregor (1936) CLR 50, 62. But there is another question about costs on which the parties wish to make submissions to this Court. It therefore seems preferable to reserve the costs both of the action and counterclaim and of the appeal until written submissions on all of these matters are provided to the Court by the parties.

[43] THOMAS JA & CULLINANE J: This case arises out of a sale of goods (PVC resin) by the respondent (Amcor) to the appellant (Metal Roofing and Cladding). It will be convenient to refer to Amcor as the vendor and to Metal Roofing and Cladding as the purchaser.

[44] The purchaser declined to pay the balance of purchase price of resin, delivered over an extended period, claiming that some of the resin did not conform to the commodity which it had ordered. The vendor sued for the full price, and the purchaser counter-claimed for damages for misrepresentation and for breach of implied conditions of reasonable fitness for a stated purpose, of merchantable quality and of other alleged conditions. The learned trial judge rejected the purchaser's contentions in both its defence and counter claim, and granted judgment against it for $589,086.50 with a declaration of the vendor's entitlement to interest.

[45] The principal issue which arises on this appeal is whether the learned judge erred in rejecting the purchaser's claim that the goods failed to correspond with their description. The purchaser also appeals against the allowance of a further claim for damages of $60,879 incurred by the vendor through its loss on re-sale of a quantity of resin which it had, by arrangement, reclaimed from the purchaser. Upon appeal we were informed that the purchaser had paid the sums ordered pursuant to the judgment, and that it no longer sought to set aside the judgment insofar as it consisted of the price of the goods. That would seem to be a correct concession in view of s 14(3) and s 54(1)(b) of the Sale of Goods Act 1896. The orders now sought on appeal are confined to orders that the vendor pay damages under the purchaser?s counter-claim ($499,497.50), that it repay the damages awarded and paid to the vendor in respect of the reclaimed goods (which with interest amounts to $80,250.69), that the vendor pay interest on the damages awarded under the counter-claim (amounting to a further $220,993.26) and for certain orders in relation to costs of the proceedings.

[46] It was submitted that the learned trial judge ought to have held that the relevant contract for the sale of PVC resin from the vendor to the purchaser was a contract for the sale of goods by description and that the goods in question did not correspond with the description. It is common ground that the goods in question consisted of PVC resin.

[47] At trial the issue of description or identity of the goods the subject of the submissions was raised by the Further Amended Defence and Counter-claim of the purchaser, which was supplied to the court on the first day of hearing. It alleged:

"PVC resin is a material which consists of at least 99.6% poly vinyl chloride, and 306 metric tonnes of the goods so delivered to it by the Plaintiff purporting to be PVC resin consisted of 96.8% poly vinyl chloride and 3.2% of an unknown substance, or 98% poly vinyl chloride and 2% poly vinyl alcohol, and as such did not constitute PVC resin.

It is a recognised industry standard that PVC resin would consist of a minimum of 99.6% poly vinyl chloride. The "Recognised Industry Standard" is found in three places. First, the term "PVC" is descriptive in that it refers to poly vinyl chloride. Second, it is customary within the Australian Manufacturing Industry which uses PVC resin that "PVC resin" would consist of a minimum of 99.9% poly vinyl chloride. Third, the Standard is set out in the American Standard for Testing Methods ("ASTM") which requires PVC to contain 100% poly vinyl chloride."

[48] The previous defence had alleged that the recognised industry standard required a minimum of 99.9 per cent poly vinyl chloride. That however was abandoned in favour of the premise that the minimum level must be 99.6 per cent.

[49] The learned trial judge's findings on this question include the following:

"The evidence supports a conclusion that it could be expected that PVC resin would contain quantities of other substances notably poly vinyl alcohol which is an agent used in the production of the resin. There was some disagreement as to the quantities which might be expected but the overall range was between 0.1 and 0.3 per cent. The evidence is far from founding a custom binding on the plaintiff. It cannot be said that the American Standard applied and there was evidence that material not containing 100% poly vinyl chloride was sold as PVC. The evidence simply does not establish that the resin supplied by the plaintiff was other than poly vinyl chloride satisfying a contractual arrangement between the parties.

In any event, the evidence founds a conclusion that there is room for differences in determining the poly vinyl chloride content of resin which is an exact task requiring the use of sophisticated equipment. It is also open to the conclusion that in the present case the chlorine microanalysis process is a more accurate method and it establishes that the resin in question was 99.5% poly vinyl chlorine. Put shortly, the defendant has not proved the breach on where (sic) it relies."

[50] It would seem that his Honour was in error in stating that the chlorine microanalysis process established the resin in question to be 99.5 per cent PVC. The only evidence supporting such a test result was in relation to PVC resin subject to manufacturer's code WS 1000. The resin in question was subject to the code S 1000. The evidence as to the results of testing the content of contaminants in the S 1000 resin that was in issue varied between 0.9 and 3.9 per cent[1], but as his Honour pointed out there is considerable variation in testing methods and their reliability. If tests other than chlorine microanalysis were to be relied on, it would be open to hold that the PVC content of the resin in question lay somewhere between 96.1 per cent and 99.26 per cent. If the figure from chlorine microanalysis (which on the evidence his Honour considered preferable) is adopted, the correct figure for the PVC content of the resin in question should probably have been between 99.1 and 99.26 per cent.

[51] It would seem unlikely that this misunderstanding affected his Honour's final conclusion concerning the subject of identification of the product, as they relate to different questions and the figure upon which his Honour acted (99.5 per cent) was in any event less than the minimum content level of 99.6 per cent upon which the purchaser relied.

[52] Five experts provided evidence in one form or another, namely Messrs Aylett, Le Hunt, Scully and Tan and Dr Truss.

[53] It will be seen that the bulk of the evidence, (including some evidence from the vendor's experts) supports the existence of an "expectation" in those dealing in the commodity of a purity level of at least 99.6 per cent although there was no consensus on what that level should be. Further there was no consensus on the more specific issue of whether material of a lesser degree of purity would be regarded by those dealing in such materials as unable to be described as PVC resin.

[54] Mr Aylett, a plastics expert and the principal expert called by the vendor, pointed out that published specifications did not specify the PVC content of the product. He acknowledged however that relevant specifications allowed a content of water and volatiles "commonly between .1 per cent and .3 per cent", and that there could be further trace levels of catalyst and reaction material up to .1 per cent. It was Mr Aylett's experience that "the remaining 99.6 per cent would commonly be expected to be composed of PVC as homopolymer" (our italics).

[55] Mr Aylett was one of the writers of a joint experts' report in which an expert retained by each party reported upon specified issues. In it Mr Aylett pointed out that there is no Australian standard which specifies PVC resin. He further noted that specifications published by Australian PVC manufacturers (and other PVC producers outside Australia) acknowledged the existence of other matter "commonly" between the levels mentioned above. This was followed by his acknowledgment that 99.6 per cent was a common expectation. Counsel for the purchaser sought to attach particular significance to a statement made by Mr Aylett in cross-examination that if he were to purchase resin he "certainly wouldn't expect less than 99.7 per cent there for PVC resin homopolymer". It should be noted however that Mr Aylett, while conceding that this would be his expectation, maintained his thesis that there was no recognised industry standard for the term "PVC". The answer given in cross-examination was in response to a question concerning "assumptions" of small amounts of residual elements. It is desirable to state his full answer:

"Yes, when we use the word "assumption" or what was normal or what is standard or what is expected, [it] is based upon what people are used to obtaining or what they believe they are obtaining so I am not disagreeing with the point, but if I was to purchase resin I would expect it to contain, as I have stated, .1 per cent of catalysts and other additives and up to 3 per cent water in the resin and as the final product, yes, I certainly wouldn't expect less than 99.7 per cent there for PVC resin homopolymer".

That in our view was quite consistent with his statement in the joint experts' report. Mr Aylett further stated that he would still call the product PVC if it contained considerably greater levels of contaminant than 0.3 per cent. Under cross-examination he said he would still call it PVC even if contaminants of 3.2 per cent were present.

[56] Mr Le Hunt, a consulting engineer, was engaged on behalf of the purchaser. Mr Le Hunt was co-author with Mr Aylett of the joint experts' report referred to above. Twenty-three issues were submitted to them, and in the result they were able to reach only qualified and limited agreement on two. Relevantly to the present issue, the following proposition was submitted to them:

The pipe resin supplied by Amcor to Pipe Makers did not constitute PVC resin as that term is understood within the Australian industry".

They then proceeded to consider the following "sub-issues":

"(A1) Whether there is a recognised industry standard for the term "PVC".

(A3) The effect of the Australian Standard test method AS1462.18 for "Determination of PVC and Titanium Dioxide Content" and results produced thereunder".

[57] Mr Le Hunt considered that anything containing less than 99.9 per cent pure PVC resin failed to qualify as PVC resin. He wrote:

In the PVC pipe industry, the conventional assumption is that the PVC resins received in the processing plants consists (sic) of 100% PVC and only PVC as a homopolymer.

In practice, there will be some residual traces of catalyst material, but these would be expected to be at a level less than 0.1% of the total content.

The convention that "PVC" means 100% PVC is apparent from a number of sources. First, formulations (that is the formula of different compounds used to create the end product) are usually expressed in terms of parts per 100 parts of PVC, with the assumption that the PVC is 100% pure. If that was not the case, the formulations would be inexact and uncertain.

Secondly, this convention appears in standard reference text such as Matthews "Vinyl and Applied Polymers" [Butterworths UK 1st ed 1972 Vol 2 at p148] where typical formulations list PVC as being 100% pure PVC.

Further, the Australian manufacturers of PVC, ICI and Auseon, recognise this convention and use the term "PVC" to mean 100% PVC.

This convention is also apparent from the Australian Standards. For example, in the case of sewer pipes, Australian Standard AS1260-1984 Part 1 requires that the PVC content shall be not less than 83.3%. AS1260 further requires that the total additives to the PVC shall not exceed 20 parts per 100 parts of PVC. These regulatory requirements imply that the PVC being used is 100% pure.

Finally this convention is recognised internationally. For example, there is a basic assumption contained in the Standard Specification for Poly (Vinyl Chloride) Resins set by the American Society for Testing and Materials that PVC is to be 100% pure [1993 Annual Book of ASTM Standard Volume 08.01]".

The differences in the approaches of Mr Aylett and Mr Le Hunt can be seen from the extracts which have already been quoted above. These differences were never reconciled, and remained for the court to resolve.

[58] Mr Scully was retained on behalf of the vendor. He was asked by his client to answer claims by the purchaser "that the PVC content in the resin supplied was low". However he also addressed the question "What is PVC?", opining that "PVC, or poly vinyl chloride, is a thermoplastic material ie it is softened by the application of heat, and when it has been adequately heated, it can be shaped into a product by the application of pressure". He pointed out that PVC resin is by itself useless and that it must be combined with certain additives depending on the end product being made. Those comments are relevant although they do not directly address the question when the product might be said to lose its identity through having an excess amount of extraneous material in it. He pointed out that PVC resin is made in batches, that there is a slight variation from one batch to another and that in practice there is considerable blending of batches throughout the manufacturing process. He stated:

"Most of the additives used to make PVC resin (excluding most of the water), or their degradation products, remain in the PVC resin. For suspension polymers, this means that commercially pure PVC resins contain at least 0.2% - 0.3% of "non-PVC". Included in this is up to 0.1% residual water. Some Japanese technology requires the use of much higher levels of additives to make PVC resin".

Mr Scully later discussed the purchaser's allegation that the "PVC content of resin supplied was low" (the allegation in the defence as it then stood was that there should be 99.9 per cent PVC in PVC resin). He pointed out the difference between Japanese technologies and European and American technologies and that different "grades" of PVC could be obtained. In cross-examination Mr Scully was directed to the question of terminology for what could properly be described as PVC resin. The following question and answer sufficiently records his view on this issue:

"Q: If you came across a product that claimed or called itself PVC, and it had ten times that amount, perhaps 3 per cent non-PVC material, would that not be PVC, in your view?

A: It would still be PVC because essentially the majority of it is PVC. 98, 97 per cent is still essentially PVC. Its purity would be different. It would be less. You are really asking how would I respond to that sort of proposition. It would depend very much on who had made it. Manufacturers not just of PVC but of any polymer are required to use certain things to achieve the end result they are aiming for. They don't all use the same technology. They may end up at the same end result by way of different groups, so I would be surprised at 2 per cent and I would probably question it a bit. If I believed the manufacturer were a reputable manufacturer and it was obvious why it would be there, I would accept it as PVC".

[59] Mr Tan's opinion is contained in the form of a summary of a discussion with him on 31 March 1998 and was received through Mr Scully. The relevant entry is as follows:

"Mr Tan was asked what was the purity of PVC suspension resin. He replied that he would expect to see the following impurities in the resin: 0.3% volatiles plus 0.1% for granulating agent and other necessary materials. That is about 0.5% non-PVC".

The language is of general expectation rather than of definition of a product, and, significantly in our view, it reveals an expectation lower than the level at which the purchaser alleged that the product ceased to conform to its description.

[60] Dr Truss was engaged on behalf of the purchaser. He stated:

"I have no doubt that from an academic or commercial view, the term "PVC" means what it says. "PVC" may contain a small amount of chemicals which arise from the production of the resin, particularly from PVA which may be added to control particle size during production. However, I would expect PVC purchased in Australia to mean a product which is at least 99.9% pure PVC".

With respect to the sale of commodities of poorer quality he said:

"In Australia, prime resin is geared at specific market outlets like the pipe manufacturing industry. Offspec resin is sometimes produced. This is resin which does not comply or fall within the normal specifications of the manufacturer [and] is marketed as such. It may be possible to use Offspec resin, but it depends upon the purpose for which it is intended to be used. In my experience Australian manufacturers would always point out the fact that the resin was "off spec" and this normally resulted in an adjustment to the purchase price".

"If PVC contained an unexpected impurity such as PVA, I believe:

(a) any amount of PVA in excess of 0.1% would make a product impure to such an extent that it would not be "PVC" as recognised in the Australian industry;

Other evidence suggests that PVA (poly vinyl alcohol) is a contaminant that causes more acute problems than other contaminants.

[61] To complete this resume of evidence on this question, some reference may be made to the evidence of Mr Listaken, a person with some experience in the purchase and selling of PVC resin. The main thrust of his evidence was that if PVC resin contained a contaminant of between 2 per cent and 3 per cent as was alleged in this case, in his experience it would be discounted in the market place by about 20 to 25 per cent. His evidence does not in our view support the existence of any cutoff point at which the product would cease to be identified as PVC resin. When asked how he would classify a PVC resin which contained poly vinyl alcohol he responded that it would be "off specification" meaning that "it doesn't conform to the manufacturer's specification for prime virgin resin". He considered that if able to be sold it would be sold at a discount. He indicated that when his company imported PVC resin, "we just rely on data sheets from the manufacturer and the resin should conform to those specifications". He also identified resin which failed to conform to manufacturer's specifications as "off-spec resin". He believed that the market in Australia buys off-spec resin from all around the world, and that there are a variety of manufacturers all of whom produce different resins that might have different contaminants. Mr Listaken's evidence is by no means conclusive but in our view it tends to suggest that there is no clear dividing point or terminology such as that advanced on behalf of the purchaser.

[62] In reviewing the evidence on this question it would seem that terms such as contaminant, contaminants, impurities, additives and extraneous matter have been used indiscriminately in referring to the non-PVC substances that are invariably and necessarily introduced into PVC resin in the course of its manufacture. The concept of "pure PVC resin" would seem to be only a theoretical one. Some difficulty is occasioned by the use of some witnesses of the term "off-spec" resin, given that there is no Australian standard or known general specification for the substance. It would seem for example that Dr Truss and Mr Listaken, who are witnesses upon whose evidence the purchaser places some reliance, have used that term with their own quite different meanings. There are of course manufacturer's specifications, and in instances where they deal with the question of purity and are produced to the purchaser they would be expected to govern the contract or at least the "expectation" of the parties in those particular instances. In the present instance the specification with respect to the Chinese manufactured PVC resin gave certain details including details described as "K factors", viscosity, polymerising strength, apparent density, volatile matter, plasticiser absorbing, residual VCM PPM max, particle distribution and dink value. However, as we read it the specification does not deal as such with the question of the overall level of non-PVC and was silent on the question of the overall level of contaminants.

[63] The question is whether the general expectations mentioned by the above witnesses require a finding that resin containing other substances exceeding .4 per cent may not properly be described as PVC resin. The evidence is capable of supporting the view that a purchaser would generally be surprised to receive anything less than 99.6 per cent purity, but that really deals with a different question. There appears to have been very little cross-examination to illuminate the issue upon which the purchaser primarily bases the present appeal. The principal argument of the purchaser is that the majority concessions in relation to "expectation" of a level of at least 99.6 per cent of PVC resin were such as to require a finding that such a purity level was an essential part of the description of PVC resin.

[64] The case was not one of the sale of PVC resin by trade name. The terms S1000 and WS1000 appear to have been manufacturer's codes. The evidence establishes that PVC is a polymer and that in its production as PVC there is inevitably produced a certain level of contaminant. In discussing this issue the learned trial judge raised the question whether there was a "custom binding on the plaintiff" (vendor) as to the level of purity asserted by the defendant purchaser. His Honour's reference to custom seems to have been derived from the purchaser's argument, mentioned by his Honour shortly before making that observation, that "it was customary within the Australian manufacturing industry which uses PVC resin that 'PVC resin' would consist of a minimum of 99.9 per cent poly vinyl chloride". We take his Honour's comment to be a shorthand reference to the need for evidence that in commercial language a product with less purity than 99.6 per cent failed to meet the description of PVC resin.

[65] It may be noted in passing that the purchaser's cause is not advanced by the discrepancy of view between its own experts, some of whom claim that the appropriate level is 99.9 per cent while others claim less. Whilst this point is by no means conclusive, it tends to go against the existence of a sound commercial understanding upon the question.

[66] Under s 16 of the Sale of Goods Act 1896 the term "sale of goods by description" has been regarded as being concerned with the question of identity, and, so far as specific goods are concerned, to refer to a sale of goods as belonging to a particular kind, class or species[2]. In Christopher Hill Ltd v Ashington Piggeries Ltd[3] Lord Diplock stated:

"It is open to the parties to use a description as broad or narrow as they choose. But ultimately the test is whether the buyer could fairly and reasonably refuse to accept the physical goods proffered to him on the ground that their failure to correspond with that part of what was said about them in the contract makes them goods of a different kind from those he had agreed to buy"[4].

[67] Issues of this kind have been considered in numerous cases including Nichol v Godts[5] where the test was stated as "the general usage of persons dealing in this particular article", and Josling v Kingsford[6] where the test for the jury was "that which in commercial language might properly be said to come under the denomination of oxalic acid". In Sopers v Johnston[7] there was a sale of bottles of drink described as "cordials". It was found that the beverages, although labelled "cordials", contained no sugar. It was held that the word "cordial" implies that the beverage contains a substantial percentage of sugar and that accordingly there had been a breach of the warranty implied by s 13 of the Sale of Goods Act 1983 (UK). It is doubtful if the issue in that case would be decided the same way today when sugar substitutes are well recognised, but this is merely to underline that the question is one of fact to be decided upon the evidence given in the particular case.

[68] Despite the preponderance of evidence supporting a general expectation of a high level of purity, we do not think that the evidence as a whole requires the finding which the purchaser now seeks from this court. At trial the contest hardly proceeded beyond the statements by Mr Le Hunt that 99.9 per cent was necessary and that anything less was not PVC; and the contrary statements of Mr Aylett and Mr Scully that despite expectations of high levels of purity, there was no such recognised limitation upon the description of the product. Even more to the point, whilst the learned trial judge did not see fit to express either a general preference for one of these witnesses over the other, his Honour expressly held that "the evidence simply does not establish that the resin supplied by the plaintiff was other than poly vinyl chloride satisfying a contractual arrangement between the parties". This would seem to be an implied rejection of Mr Le Hunt's evidence on this question.

[69] It was submitted for the purchaser that there was evidence that the market would discount the price by at least 20 to 25 per cent in respect of a product the quality of that supplied by the vendor. That however introduces the subject of quality which is quite different from the question under consideration. A careful purchaser concerned with the subject of quality might stipulate the quality desired and pay for it according to market price. In any event the sale of the relevant resin, which was known to be manufactured in China, was made after provision to the purchaser of the manufacturer's specifications and the supply of a relatively substantial sample (100 kilograms). The learned trial judge referred to the fact that the purchaser conducted a test before placing orders for substantial quantities. There was and is no suggestion by the purchaser that it failed to conform with the sample. The learned trial judge held that the purchaser ordered the PVC resin by that description relying on its own judgment based, among other things, on the manufacturer's specification sheets and of the test it conducted. No case was established that these specifications were erroneous or misleading. His Honour also noted that certificates of analysis were provided from time to time; that the purchaser had made its own inquiries and exercised its own judgment; and that there was an absence of any complaint about the resin until the issue of payment became pressing.

[70] The provision of a sample and conformity of the bulk with it would not defeat the purchaser's claim if the evidence requires the finding that the relevant goods did not conform to their description. The sequence of events however seems a little odd if there is such a well-known identification of the product as the purchaser submits. The evidence demonstrates that there is no international standard in respect of the product and indeed that different expectations and manufacturing processes exist in different countries. The context of this transaction is a sale within Australia of a product that the parties knew to be manufactured in China, which followed earlier contracts for sale of PVC resin known to be manufactured in Korea, and which followed provision of the manufacturer's specifications and of a sample.

[71] Looking at the evidence as a whole we see no reason to think that his Honour erred in holding that it was not established that the resin failed to satisfy its description. Although the error which his Honour made with respect to the chlorine microanalysis test is a matter of some concern, as mentioned earlier it does not falsify his Honour's conclusion on the question which was litigated concerning what was generally understood by the term "poly vinyl chloride". The purchaser does not seek a retrial, and its submissions are confined to persuading this court to substitute a different finding on this issue in place of that of the learned trial judge. The absence of further discussion by his Honour of the divergent approaches of the expert witnesses is a factor which might tend to cancel out any advantage that the learned trial judge had in observing the witnesses and considering the acceptability of the opinions expressed[8]. But notwithstanding this, having reviewed the whole of the evidence we are not constrained to reach a different conclusion from that reached by his Honour.

[72] Indeed, the variation between the evidence of every witness who addressed this issue suggests to us that at the material time there was no general or uniform understanding or expectation of the tolerable level of non-PVC substances, let alone of the level at which resin could no longer be properly described as PVC.

[73] It follows that there is no basis upon which the counter-claim can succeed.

[74] It is therefore unnecessary to assess the damages, as no relevant breach has been established. We have had the opportunity of reading the reasons of McPherson JA who would set aside the learned trial judge's findings and award damages on the counter-claim. Were a reassessment necessary, we should indicate that we would not be prepared to make a finding that the PVC resin supplied by the vendor to the purchaser in 1995 contained a poly vinyl alcohol contaminant of about 2 per cent, as suggested by one analysis that was conducted in May 1996. There is too much conflicting evidence for this court to essay satisfactorily the average level of non-PVC content. Another test produced an average result of one percent, and we apprehend that on such a finding the damages would be assessed at a significantly different level to those based on a finding of 2 per cent impurity. If damages need to be assessed it is preferable that the judge who had the advantage of seeing the witnesses should make a finding as to what evidence is accepted, rejected or preferred on these questions. There is the further possibility that if a deal of uncertainty surrounds the best finding that can be made (as is likely in a case where only a few disparate samples are taken of substantial material delivered over an extended period) some discounting might be appropriate.

[75] If, contrary to our view, the purchaser is entitled to judgment on its counter-claim, the question of damages thereon ought to be referred back to the learned trial judge with a direction to assess such damages and enter such judgment as is just.

[76] A further issue is whether the vendor's general terms and conditions form part of the contract between the parties. This is relevant to the rate of interest and the form of order as to costs. His Honour found that the General Terms and Conditions become part of the contractual arrangements between the parties as a result of a course of conduct. We think that this finding was justified by the evidence and in particular the evidence of the significant period over which the vendor supplied the defendant with PVC and the evidence of Mr Speight the general manager of the plaintiff which is to the effect that the General Terms and Conditions are printed on the back of every invoice, delivery docket and statement generated by the defendant in the ordinary course of business.

[77] We also agree with McPherson JA that the General Terms and Conditions were expressly incorporated into the terms upon which the parties dealt by the purchaser's acceptance of those Terms and Conditions when making application for credit. The purchaser agreed to accept the General Terms and Conditions at that time and as McPherson JA says, they became binding on it at the latest, if not before, when the first supply of PVC was ordered and satisfied. We also agree with McPherson JA that the General Terms and Conditions were sufficiently identified.

Vendor's damages from resale of reclaimed resin

[78] The purchaser also appeals against the finding that it is liable in damages of $60,879 for damages arising from the re-taking of 144.95 metric tonnes reclaimed from the purchaser after delivery. The basis of the claim is contained in paragraphs 14, 15, 15A and 15B of the Amended Statement of Claim. The allegation is that pursuant to its right to do so under clauses K, L and Q of its general terms and conditions, and with the purchaser's consent, the vendor recovered that quantity of resin and issued credit notes in favour of the purchaser in the sum of $246,655. However on re-sale the vendor was only able to recover $184,086.50. Accordingly $60,879 was claimed as damages. We agree generally with what McPherson JA has written on this issue. The point which the purchaser now seeks to pursue was not raised on the pleadings, and the issue does not seem to have been seriously pressed before the learned trial judge. It seems to us that this claim for damages by the vendor was posited upon an original entitlement to payment for the sale and delivery of this particular resin. The evidence of Mr Speight explains the return of the goods as being a consequence of the vendor's acceptance of the purchaser's "invitation" of 7 November 1995 to repossess the resin and to issue a credit for it. The letter in question deals with the then dispute between the parties concerning weight discrepancies. In the result, the purchaser failed to establish any legitimate grievance on this basis or indeed upon any of the bases raised by it to show that the vendor was in breach of its obligations. It would therefore seem that on the ultimate findings that had been made and upheld the purchaser was obliged to pay the price and it has been given the benefit of the amount obtained by the vendor on resale. It might also be possible to see the purchaser's conduct as an unjustified non-acceptance of the goods. In the end the fact that the parties were not ad idem as to the entitlement of the vendor to sue for any loss on resale does not matter.

[1] Testing at the WA Chemistry Centre revealed contaminants in such resin at a level of 3.2 per cent. Testing by European Vinyl Corporation revealed 2 per cent; evidence in an analytical report by Probe Analytical suggests (by hearsay) a chlorine microanalysis test showing average contaminants of .9 per cent, and an Australian Standard test showing average contaminants of 3.9 per cent.

[2] Compare Salmond J in Taylor v Combined Buyers Ltd [1923] NZGazLawRp 133; [1924] NZLR 627, 639; Scott LJ in Couchman v Hill [1947] KB 554, 559.

[3] [1972] AC 441.

[6] [1863] EngR 213; (1863) 13 CB (NS) 447 at 451.

[7] [1944] 2 All ER 586.

[8] State Rail Authority of New South Wales v Earthline Constructions Pty Ltd (In Liq.) [1999] HCA 3; (1999) 73 ALJR 306 at 330 par 91 per Kirby J; Ahmedi v Ahmedi (1991) 23 NSWLR 288, 290, 299; Walton-Taylor v Wilson (NSWCA, CA 40300 of 1995, 26 March 1998, [1998] NSWSC 79, BC9800882 per Priestley JA at p 30; Handley and Stein JJA concurring).