KOLINKER INDUSTRIAL EQUIPMENT LTD v. LONGHILL INDUSTRIES LTD & ANOTHER

DCCJ5052/2002

IN THE DISTRICT COURT OF THE

HONG KONG SPECIAL ADMINISTRATIVE REGION

CIVIL ACTION NO. 5052 OF 2002

---------------------

-------------------

_____________________

Judgment

_____________________

1. The plaintiff is a Hong Kong company, in the business of making semiconductor equipment. One of the things which it makes is called a "package"; that is a little sealed box, with contacts under its four corners, containing a crystal. The package must be hermetically sealed by welding on a metal lid. This is done by an automatic welding machine called a "seam sealer". Before the package gets to the seam sealer, the lid must be precisely placed and this may be done by tack-welding it in position, using another automatic machine called a "tacker".

2. The defendant is also a Hong Kong company, in the business of selling and servicing semiconductor equipment. In 1999, it had sold the plaintiff a seam sealer manufactured by the third party ("SSEC"), which is an American corporation. The plaintiff used that seam sealer to make packages in its Shenzhen factory. At that stage, the lids were being placed on the packages manually, before sealing. The plaintiff decided to buy a tacker, and by a purchase order dated 8 May 2000 it bought from the defendant the Automatic Matrix Lid Tacking System, model no. 2300e, also manufactured by SSEC; and it is this machine which is the tacker with which this case is concerned.

3. The plaintiff's case is that the tacker was of a particular pleaded description which related to the precision accuracy of placement of the lids on the packages, and speed of operation. In particular it had to place the lids to within 0.002 inches, and tack over 500 packages per hour. It was an implied condition of the agreement between them that the tacker should conform to that description. Further the defendant entered into the agreement in the course of its business and knew that the plaintiff required the tacker for use in its business, and accordingly there were implied into the agreement conditions that the tacker should be fit for the purpose for which it was required, and of merchantable quality. The tacker was delivered in November 2000, and the plaintiff paid 90% of the price plus incidental expenses. But it never worked properly. It misaligned the lids, giving an average 7.8% error rate; there was a gross leak or airtight failure in the sealing function with an average 6.7% error rate; and the tacker only tacked 350 packages per hour as against the 500 pieces per hour described. It was not reasonably fit for the purpose for which it was required, nor was it of merchantable quality. The plaintiff was entitled to reject the tacker and did so on 23 May 2002; and now it claims its money back. There was also a claim for consequential damages, but there has been no evidence to support such a claim, and when this was pointed out in the course of the trial counsel for the plaintiff agreed that it would not be proceeded with.

4. One leg of the defendant's case was originally that it had merely sold as agent of SSEC but this was abandoned at trial. The defendant however says that the plaintiff was subject to SSEC's terms of warranty which were incorporated into the agreement by way of previous dealings. One of these terms or alternatively a term implied into the agreement provided that the plaintiff had to return the tacker to SSEC to invoke its right of rejection. It failed to do so despite repeated requests, and accordingly it is not entitled to reject the tacker, even if the tacker is defective.

5. So far as the tacker itself is concerned, the defendant's case is that it did not warrant or represent any of the alleged description; or the purchase order never specified the alleged description; or the alleged performance description was invalid or inapplicable because, in effect, the performance specified would only be obtained if the plaintiff installed a "bowl feeder" for use with the tacker. The plaintiff had originally intended to use a bowl feeder, but later changed its mind; and it knew that a bowl feeder was necessary for the efficient working of the tacker.

6. In any event the defendant says that there was no inadequacy in the lid placement accuracy or the speed of the tacker; it performed normally and it was of merchantable quality. Alternatively any inadequacy arose out of the plaintiff's wrongful failure to install a bowl feeder. Any gross leak error arose out of the poor quality of the plaintiff's materials, or from the seam sealer and not the tacker. In any event, if the defendant was in breach of the agreement, the breach was so slight as not to entitle the plaintiff to reject the tacker; the rejection was unjustified and the plaintiff is not entitled to claim its money back.

7. As will be seen, the tacker, once delivered, never worked properly and was never used in production. Even the defendant's director had to admit in the course of cross-examination that "you can say that the machine is comparable to junk" although of course his evidence was the various problems with it were beyond the defendant's control.

The Tacker

8. The tacker was brought into court. It is a machine about two or three feet wide and about two feet high which will sit on a table or workbench. In operation it is connected to and controlled a personal computer; the operating software is supplied by SSEC.

9. The tacker works in this way. The packages are placed in cavities in a "matrix tray" which sits on a turntable and is located precisely by means of two vertical pins which fit into location holes machined into the bottom of the tray. There are 80 or 90 cavities in a matrix tray and they are made so that the packages fit snugly in them and are held in place by little magnets beneath each one. The lids are stacked one on top of the other in a "lid bin tool" which is a steel block with 20 vertical bins in it, machined to the size of the lids which, in the plaintiff's process were either 7x5 or 6 x 3.5 millimetres. The lid bin tool sits to one side of the turntable. A "picker" which is a suction nozzle picks up the top lid from a bin and carries it past a digital camera which is called "vision on lid"; the function of this is to align the lid so that it can be placed on the package. The picker then carries the lid past another camera called "vision on package" whose function is to identify the position of the package so that the picker can place the lid on it. When the picker has done that the vacuum suction goes off and a double electrode moves to the lid and places a tack-weld on both the long sides. The process is then repeated until all the lids in the matrix tray have been tacked. Another tray is placed manually on the turntable and the process is repeated. The trays may be loaded automatically by a cassette but that was not provided for on the tacker in this case.

10. When the bins in the lid bin tray are empty they are also refilled manually. In place of the lid bin feeder fitted to the plaintiff's tacker, there may be fitted a "lid feeder" or "cassette"; the salient feature of this is that it pushes each lid up the lid to the top of the bin in turn, so that the picker does not have to go further down into the bin as the lids are progressively taken away. Alternatively the bowl feeder referred to in the pleadings may be fitted. The salient feature of this is that it propels each lid in turn along a track by vibration, to a set position where it lies without vibrating, so that the picker returns each time to that position to pick up a fresh lid.

The Documents - before delivery

11. It is not in dispute that the defendant supplied the plaintiff with SSEC's catalogue for the machine and a set of throughput calculations for it and that a Mr Richardson of SSEC, at a meeting attended by representatives of the parties on 11 November 1999 wrote out a table to show the cost of the machine and the speed of production of sealed packages in conjunction with the seam sealer. In brief, according to the throughput calculations, the production using the tacker in conjunction with the seam sealer was supposed to be 558 seconds per tray, or 516 packages per hour with the 6 x 3.5mm. packages, or 435 seconds per tray or 529 packages per hour with the 7 x 5mm. packages. The catalogue contained certain assumptions one of which was "Lid Feeding from Cassettes. Bowl Feeder Increases Speed Further". The throughput calculations contained a note which reads

"Note: Time does not include operator intervention for changing cassettes and lid feeders. Bowl Feeder for lids are available."

12. Mr Richardson's note compared the prices and the speeds of various methods of preparing the packages for the seam sealer. With "lid pick and place" at a price of US$40,000.00 the production speed was estimated at 350 units per hour. With that plus a tacker, at a price of US$65,000.00 the estimated speed was 500+ units per hour. These figures were confirmed in an e-mail from SSEC to the plaintiff, and copied to the defendant on the following day.

13. On 17 November the defendant sent the plaintiff a quotation for the tacker at a price of US$62,774.00. Various specifications and optional extras were included. Along with this quotation, SSEC sent to the defendant, who transmitted it to the plaintiff, a fax message which contained some significant information, as follows:

"SSEC's strategy is to provide the most accurate lid placement with a highly flexible system, hence we use vision on lid and vision on package. This eliminates the expense of a bowl feeder and enables rapid size change over in minutes.

Please see the attached document "Small Package Lid Tacking System". In this, we are sealing 80 packages at 6.45 seconds per package or 558 packages per hour. This is with cassette automation which takes well less than 10 seconds to change the tray.

With this data they can calculate how long it takes operator to remove one tray and replace it with another tray and press the start button. Maybe this takes 15 seconds, for instance, which does not affect these numbers a whole lot".

14. The attached document was the SSEC catalogue.

15. The parties negotiated and the defendant sent the plaintiff a revised quotation no. Q111701 R1 with a 5% discount on 2 March 2000. The covering letter said that the defendant offered specially a free-of-charge design and AutoCad drawing of Lid Bin Tooling and Matrix Tooling, and these elements were included in the quotation. The price of the tacker was now US$59,635.00. Included in the specifications was that pleaded, namely that the machine had a placement accuracy of ±.002". Also included was the following specification on which the defendant relies

"d) Automatic Lid Feeding

^. Accepts Standard Lid Cassettes from Lid Suppliers for Reduced Lid Handling

^. Multiple Lid Bins with Lid Tooling area of Standard 2.5" x 6" up to 3" x 6"

^. Light Tower System Provides Notification on Lid Materials by Bin"

16. On 31 March 2000 the defendant sent an e-mail to SSEC to the effect that the plaintiff was considering using a bowl feeder rather than a lid bin tool, asking for information about the bowl feeder for the plaintiff and passing on the plaintiff's own design for one of the matrix trays (for the 7mm x 5 mm packages) for the approval of SSEC. On 13 April 2000 the defendant sent the plaintiff a layout for the bowl feeder and approval of the matrix tray as "acceptable for seam sealing". It added that "SSEC would caution you about not having excessive clearance between the packages and the tray cavities.". This was followed on 20 April by a message from the defendant to SSEC that the plaintiff worried that the bowl feeder might not be so smooth in early stages and were thinking of having a stand-by lid bin tool, mentioning the fact that drawings of the lid bin tool were to be provided free of charge and asking for a drawing. SSEC sent such a drawing to the defendant on the same day.

17. The plaintiff then sent the defendant its purchase order dated 8 May 2000. It said that the detailed specifications should be as per quotation no. Q111701 R1 and added

"INCLUDED FREE INSTALLATION AND TRAINING AT OUR SHENZHEN FACTORY.

WARRANTY: 1 YEAR (FREE ON SITE MAINTENANCE.)"

18. There followed a message from SSEC to the defendant asking if the plaintiff "still wanted a bowl feeder". The defendant replied that the plaintiff had decided to go for the lid bin tool first in order to avoid any possible problems during start-up, and asked for the free-of-charge lid bin tool drawings for both sizes of package for the plaintiff's use.

19. This was followed by the letter of credit, by which the defendant was to be paid 90% of the cost up front. Among the required documents was the beneficiary's statement guaranteeing that if there were any design problems on the tacker, any necessary replacement parts and related transport charge would be on account of the seller. It is not in dispute that 90% of the price was paid to the defendant.

The Documents - after delivery

20. The various documents passed between the Shenzhen factory and the defendant, or the plaintiff and the defendant, or the defendant and SSEC but there is no dispute that what is in them is factually correct. The tacker was installed on 20 November 2000 at the plaintiff's factory in Shenzhen. The complaints started on with an e-mail from Donny Shum of the plaintiff dared 12 December 2000 complaining that the machine still did not work, apparently because of a software bug. On 18 December there was a complaint that the tray was shaking and the picker nozzle was broken. On 1 February 2001, the complaint was that the picker was picking up two lids at a time, and that many lids were incorrectly welded. The same complaint, with more detail, appears on 8 February 2001. SSEC came into the picture, asking technical questions in the hope of assisting. On 19 March 2001 the complaint was that the welding head shifts the tray so that it rotates and the lid is placed slightly out of position. On 3 April, it appears that following a visit by Mr Lai of the defendant (its engineer and expert witness) with a new turntable, that problem was fixed but still the double lid picking was experienced. Out of 90 packages, 42% were not welded and of the 58% welded, 35% were welded incorrectly.

21. On 3 April 2001 Mr Arthur Lee, the technical director of the plaintiff wrote to SSEC complaining and saying that he was considering buying a Japanese machine for another factory of the plaintiff, and also seriously considering sending the tacker back. Mr Richardson of SSEC wrote on 4 April suggesting that the machine be returned to SSEC to be fixed or replaced with a new, more expensive high-speed system. He also raised the question of whether it had been the plaintiff's original intention to use a bowl feeder. The plaintiff replied that the bowl feeder was never used, so this was not causing problems. As to the suggestion of returning the tacker, the plaintiff said that the customs procedures for returning it from the Mainland to Hong Kong were long and complicated, so a new machine should be shipped out before the old one was sent back. That SSEC would not accept, however; they wanted the old tacker back first. This met with a response from Mr Lee who was obviously most frustrated, stressing that he had bought the tacker, which had never worked, and it was up to SSEC to make it work or replace it.

22. Then on 20 April 2001 Mr Lai of the defendant went to Shenzhen and tried a new computer programme. By now the machine managed to weld all but 3-5 lids in a 90-package tray; but the picker was still picking up two lids at a time. The engineers looked into the question of whether a magnetic field was involved. On 2 May the Shenzhen factory manager thought that magnetisation was the problem, and suggested a change to a vibration tray or other non-stacking structure. Ultimately, however on 3 May 2001 it was reported that whether the lids were magnetised or de-magnetised the number of packages with incorrect lid placing was similar.

23. Meanwhile, on 23 April, there was a report of 85 lids welded out of 90, but 21 of those welded were welded incorrectly. SSEC suggested that humidity was making the lids stick and suggested a vibratory bowl. Following another message expressing Mr Lee's frustrations, SSEC said that Mr Fok of the defendant thought that shallower bins were needed, so had sent to the defendant a drawing for a new lid bin tool with shallower bins. Mr Lee was not happy with this; he said that irrespective of lid sticking, there was the problem of misalignment. Mr Richardson of SSEC replied that that was a matter of training. Then they tried a new nozzle, which vibrated to shake off any extra lid. Correspondence, and attempts to fix the various problems went on and in early July 2001 one Kevin Yang, an employee of SSEC from Taiwan was sent to the factory but he could not fix the machine either. He suggested more training. But even by this time only 85 out of 90 packages in a tray were being welded; a 6.66% failure rate, as the factory manager pointed out. This was insufficient for the plaintiff.

24. Attempts to fix the various problems went on. By 5 January 2002 the plaintiff was complaining, not only that a 95% yield was unacceptable but that the tacker was only tacking 180 to 240 packages per hour, as against the speed claimed by the manufacturer. Then on 21 January 2002 personnel from SSEC, with Mr Lai of the defendant went to the factory. They reported that after certain changes had been made, on test the tacker was producing a 2% lid position poor rate; a 20% poor rate on gross leakage test after seam sealing; and a welding speed of 200-360 pieces per hour. The plaintiff's position was that was quite unacceptable; it wanted a pass rate of 100% on fine leakage testing as well as improved speed. Things got no better. By 19 March 2002 the factory manager was reporting as follows:

1. Position pass rate is 98% now, but need mass production, stably achieve 98%.

2. Gross leakage pass rate is 80%, need gross leakage and fine leakage to 99.9%

25. After that SSEC developed a new sealing schedule, and the plaintiff agreed to try it. However, on 30 April 2002 we see the results obtained by the plaintiff; according to the e-mail, 8 pieces out of 90 were found misaligned, i.e. 8.8% no good; and when the remaining 82 pieces were sealed, 5 out of the 82, or 6.9% failed the gross leak test. The response of SSEC was to ask for a matrix tray to be sent, so that SSEC could try out the system, using a matrix tray. However the plaintiff replied on 30 May 2002 purporting to return the tacker and asking for its money back.

The Joint Inspection

26. The plaintiff's independent expert, Dr. Li and the defendant's in-house expert, Mr Lai, inspected the tacker on 6 February 2003 for the purposes of this litigation. The inspection was carried out at the plaintiff's Tsuen Wan factory. Two test runs were done, the first one using a matrix tray containing 90 packages and the second a matrix tray containing 81 packages. The first test produced a speed result of 332 packages per hour and the second 339 packages per hour. After visual inspection Dr Li claimed for the first tray 8 passes and 82 fails; Mr Lai claimed 56 passes and 34 fails. The machine was reset and re-calibrated. This time Mr Lai claimed that all the packages tacked (four were not) passed the visual inspection but Dr Li said there were at least 11 fails. After microscopic examination Dr Li claimed 15 fails and Mr Lai agreed 13 fails.

27. Mr Lai in his report said that the tray used had an oversized location hole and this could have caused or contributed to the failure rate. On the second try, however, the hole was shimmed up to provide a tight fit. In evidence Dr Li emphasised that the tray used was acceptable to Mr Lai.

The Issues

28. It is clear from the above that the tacker was never of any practical use to the plaintiff. It never did what the plaintiff expected of it. The description of it as "junk" is accurate in practical terms. The issues are, however, whether there were the implied conditions which the plaintiff pleads, and whether, if so, the defendant breached them. There is also the question of whether the plaintiff was entitled to reject the tacker, or required to send it back to the defendant or SSEC in terms of the SSEC warranty.

Implied Condition of Description

29. Section 15(1) of the Sale of Goods Ordinance, Cap. 26 provides that where there is a contract for the sale of goods by description, there is an implied condition that the goods shall correspond with the description. The description pleaded relates to the placement accuracy and the speed. The defendant's case is that both the accuracy and the speed depended on the use of an automatic lid feeding device, which the plaintiff did not use. In any event although the placement accuracy was specified in the defendant's revised quotation the speed was not, and so it is not part of the contractual terms. Insofar as representations as to speed were made by Mr Richardson of SSEC it is not pleaded that they were part of the contractual terms between the plaintiff and the defendant.

30. The defendant relies in particular on the specification for "Automatic Lid Feeding" which I have set out above. In addition it relies on the references to cassettes, cartridges, lid feeder and bowl feeder which appear in the catalogue and the throughput calculations. According to the defendant's evidence, both from Mr Fok the director and Mr Lai the engineer, a bowl feeder or a lid feeder would produce both greater speed and higher accuracy. As to speed, the point is that as the number of lids in the bins of the lid bin feeder decreases, the picker has to go down progressively further into the bins, as it picks out the lids. It has to slow down as it does so. The time taken is called "seek time" or "search time". But with a bowl feeder or a cartridge that pushes the lids up, the picker goes back to the same location each time it has to pick up a new lid. As to accuracy, the bin has to be sufficiently oversize to let the lid come out smoothly without touching the sides; this can lead to inaccuracy in alignment. The cartridge feeder can have the bin machined more tightly, so that this inaccuracy will not occur. Similarly, the bowl feeder always takes the lid to exactly the same place and in the same way inaccuracy is avoided.

31. The plaintiff's witnesses, both Mr Shum and the independent expert Dr Li accept that the tacker will work faster with a bowl feeder or a cartridge feeder than with a lid bin tool. They do not accept that there are likely to be problems with accuracy in the case of the latter; what they rely on is the vision on lid camera. There is support for this in SSEC's fax message of 17 November 1999, which I have set out above; it says in effect that precisely because of the vision on lid camera the expense of a bowl feeder can be avoided.

32. The only concrete evidence as to speed which any of the witnesses advances is that of Mr Lai. While Mr Shum agreed that a cartridge feeder which pushed up the lids would work let the tacker work faster, he could not be any more definite. Mr Lai however said that the figure achieved on the experts' test of 327 or 360 lids per hour was reasonable without a bowl or cartridge feeder. Dr Li, the plaintiff's expert however took the view that although there would be an improvement with a bowl feeder, it should not be nearly twice as fast. In any event most of the time is spent on travelling and the vision system so that there should not be substantial improvement. And since the travelling time for the picker was very short, pushing up the lids would also not produce substantial improvement.

33. It has to be noted that Dr. Li is an independent expert. Mr Lai, though accepted as an expert, is "in-house". He is not independent and his evidence has to be looked at in that light.

34. It is true that the catalogue says that lids are "taken from cartridges or a bowl feeder". One of the assumptions is "lid feeding from cassettes"; this means, on the evidence, a cartridge or some other device that pushes the lids up. The bowl feeder is alleged to be faster. We do not have an estimate of the difference, if any between the bowl feeder and cartridge feeder when it comes to speed.

35. The throughput calculation mentions a bowl feeder but as an optional extra only. It and the catalogue both mention the use of cassette feeding to change the matrix trays but it does not appear that that would make much difference to the speed in any event; see the fax message from SSEC set out above.

36. Nowhere in the first quotation, or the second quotation or the purchase order (the contract documents) there any specification that the speeds estimated could only be reached with a bowl or cartridge feeder. Nor is there any evidence that the plaintiff was told this; I myself asked Mr Fok about it and he said that he himself did not tell the plaintiff that. Mr Shum said that he had asked Mr Siu of the defendant what was meant by "cassette" and was only told that it was an object on which the lids were placed. He had found nothing in the relevant part of the quotation to make him think that some kind of pushing feeder was necessary. The obvious question is, if the machine was only going to tack 360 packages per hour with a lid bin feeder against the background of documents which say that it will tack over 500, why did no one tell the plaintiff so?

37. The paragraph relating to "Automatic Lid Feeding" in the quotations seems to me to refer rather to a lid bin feeder with cassettes only as an optional extra. If one looks at the quotation as a whole one sees that the defendant was offering drawings for a lid bin feeder, free of charge, as part of the deal. I do not think that references to "lid feeder" must necessarily mean something other than the lid bin tool which was fitted. Nor do I think it matters that the plaintiff at an early stage thought about using a bowl feeder, given that the quotation itself, by including the drawings for a lid bin tool indicates that what was in the mind of the parties was that a lid bin tool was to be used. Indeed the defendant's side (including SSEC) only seems to have become interested in the question of the bowl feeder later on, when the tacker could not be made to work properly; that was in April 2001.

38. For a statement or promise to be contractual the other party must, at the time of the contract, rely on its being made good. See Benjamin's Sale of Goods, 6th Edition at 11-009. According to Mr Shum, the plaintiff was already achieving 300+ packages per hour without a tacker as well as 99% yield. There would be absolutely no point in the plaintiff buying a tacker if it could not get better speed than that, coupled with the same accuracy. I have no doubt that the plaintiff relied particularly on the figures in the throughput calculations and Mr Richardson's note. These of course came from SSEC but the defendant obviously adopted them. The defendant never said to the plaintiff, "you will only get the performance stated, if you install a cartridge or bowl feeder"; it simply went ahead and prepared the drawings for the lid bin tool, and indeed ultimately approved the lid bin tool as built by the plaintiff.

39. I do not, therefore, see that the defendant can say either that compliance with the accuracy and speed descriptions is subject to the use of a cartridge or bowl feeder, or indeed that the speed description is not part of the contract.

40. As to whether there was a breach of those conditions, there quite obviously was. As I have indicated above, the best speed the tacker ever managed seems to have been 360 packages per hour. That is the best speed the parties could get when their personnel tested it in January 2001 and it a little over the best speed the two experts could get out of it on their joint inspection on 6 February 2003; and these speeds were achieved, not in production, but with engineers gathered round the machine.

41. As to accuracy, Mr Shum said that the plaintiff needed accuracy such as to give a 99% yield. Dr Li said that for manufacturing equipment such as this the failure rate (of the whole process) should be less than 1%; in other words a 99% yield. He is an independent expert and I accept what he says on this. From all the documentation as well as the evidence in particular of Mr Shum, the tacker never came anywhere near that. It is true that there is no specific evidence that, on test, the machine did not achieve the actual accuracy of ±0.002" but it must be inferred that it cannot have done or the accuracy would have been greater.

42. The experts' test is criticised because it does not specifically show figures as to accuracy. Also the plaintiff has all along been relying, not on figures which come from the tacker alone but on the leak test results obtained after the whole process, i.e. tacking followed by seam sealing. It has been suggested that there could have been something wrong with the seam sealer. I have to say that this seems unlikely given Mr Shum's evidence that he was getting better than 99% accuracy from the seam sealer before the tacker was bought. In any event that question was never raised in the long period of attempting to get acceptable results. But I think the point is that the testing cannot be criticised. The tacker was sold in the first place to be used in conjunction with the seam sealer. If it did not work properly, then the packages would not be sealed properly by the seam sealer.

43. There are also suggestions that there might have been something wrong with the plaintiff's matrix trays. The experts got a bad one, for their test; the location hole was oversized though they managed to shim it up for their second test run. Also Mr Lai points to apparent oversize of some of the package cavities on the photographs. Again it has to be noted that SSEC approved the plaintiff's matrix trays, even if they did caution against excessive clearances; and in any event no one on the defendant's side raised the question of even looking at the matrix trays until April 2002. I do not think there is anything in this.

44. I have no doubt that the tacker did not conform to the description and that there was a breach of the implied condition.

Implied conditions of merchantable quality and fitness for purpose

45. By section 16 of the Sale of Goods Ordinance, where the seller sells goods in the course of a business, there is an implied condition that the goods supplied under the contract are of merchantable quality. Further, where the seller sells goods in the course of a business and the buyer, expressly or by implication, makes known to the seller any particular purpose for which the goods are being bought, there is an implied condition that the goods supplied under the contract are reasonably fit for that purpose, whether or not that is a purpose for which such goods are commonly supplied, except where the circumstances show that the buyer does not rely, or that it is unreasonable for him to rely, on the seller's skill or judgment.

46. Here there is no dispute that the defendant sold the tacker in the course of its business. Mr Shum says that the defendant well knew what the plaintiff wanted it for and indeed it could not have failed to know; the machine is specialised, it will only do one thing, and the whole thrust of the throughput calculations was that it was going to improve production when used with the seam sealer.

47. The defendant says that I should adopt a pass rate of 98% when considering whether the tacker is merchantable or not. This is on the basis, apparently, of the Shenzhen factory's grudging acceptance that 98% would be all right, if it could be achieved in production. But this is simply not good enough. It was coupled with a gross leakage pass rate of 80%. Dr Li says that the standard is a failure (which would mean passing both the gross and fine leakage tests) of less than 1% and I accept this. The simple fact is that the tacker was no use. It did not do what it was supposed to do, even though the parties and the manufacturer tried to make it work over a long period.

48. It is also argued that the joint test results are not conclusive to show that the machine was defective. It could have been extrinsic factors or parameter settings that led to the lid misalignment. In particular the defendant prays in aid the oversized location hole. But the oversized location hole was shimmed up to provide a tight fit for the second test. Mr Lai did mention the possibility that it might not have stayed tight; but again we come back to the question, if that was a possibility, why not raise it then and do another test? Obviously everyone wanted to get home that day; but there was nothing to stop them from coming back another day; Dr Li says that he was prepared to do that. In any event, we are not just looking at a test in 2003. We are looking at a whole lengthy history of a machine defective from the day it was installed.

49. It is also argued that the problems which occurred before the test in January 2002 should not be taken into account by the court to consider whether the plaintiff can reject the machine. All those problems had been rectified by then. The plaintiff had waived them and cannot rely on them as grounds to reject the tacker. The plaintiff refers to the case of Stocznia Gdanska SA v Latvian Shipping Co & Ors. (No. 3) 200 2 All ER (Comm) 78 for the concept of "a middle ground between acceptance of repudiation and affirmation of the contract, namely the period when the innocent party was making up its mind what to do". For my own part I cannot see any question of waiver or affirmation of the contract here. The machine never worked properly. It had a lot of problems. I do not see how there could be any question of accepting it, until it was made to work properly. To say that as each problem was solved, the buyer could no longer rely that problem is simply unrealistic.

50. Benjamin at Para. 11-033 refers to two types of test of merchantable quality, the one based on the notion that the goods must be acceptable to a reasonable buyer and the other on the notion of fitness for purpose; in effect an overlap between the statutory requirements of merchantable quality and fitness for purpose. For an illustration of the latter test the learned authors cite the opinion of Lord Wright in Cammell Laird & Co. v Manganese Bronze & Brass Ltd. [1934] AC 402 at 430:

"What subsection 2 now means by 'merchantable quality' is that the goods in the form in which they were tendered were of no use for any purpose for which such goods would normally be used and hence were not saleable under that description"

51. In this case the tacker was a very specialised precision machine. It had only one function and it had to perform that function with such precision that the product which it processed, once it had been further processed in a seam sealer, had a failure rate of less than 1%. If it could not do that it was of no use for the purpose for which it would normally be used. In effect it was "junk". It could not do that and it was therefore not of merchantable quality.

The SSEC warranty

52. The term on which the defendant relies reads:

Any equipment considered defective under the terms of this warranty must be returned immediately to SSEC, transportation charges prepaid, and will be returned to the customer, transportation charges collect.

53. This is no doubt a term of the contract between SSEC and a buyer. The defendant's case is no longer that it was merely an agent for SSEC. There is nothing in the contractual documents to suggest that this term is imported. Nor is there anything to suggest that it could be implied. Indeed I cannot see how it could be implied, given the warranty for one year's free on site maintenance. The two are inconsistent. I do not see that there is anything in this point.

54. The defendant's case has from the outset been an attempt to raise possible reasons for why the tacker did not work, which might have exonerated the seller (see in particular the lengthy witness statement of Mr Fok) followed by ingenious legal arguments, both of which have been calculated to obscure the fact that the defendant sold the plaintiff a precision machine which did not do its job. Whether it was a new machine which had not yet got the "bugs" out of it, I do not know, but that is the way it looks. In any event it could not in any sense be regarded as being of merchantable quality or fit for the purpose for which it was sold and the plaintiff had every right to reject it. The plaintiff is, accordingly entitled to judgment for repayment of the price as well as the incidental charges. Of course the plaintiff will not be entitled to keep the machine; it must let the defendant take it away if the defendant still wants it; but that is not something which need concern the court at this stage.

55. There will accordingly be judgment for the plaintiff against the defendant for the sum of US$52,771.50 or its equivalent in Hong Kong currency plus HK$4,820.25 with interest on those sums from the date of the Writ to the date of judgment at the best lending rate plus 1%, and thereafter at the judgment rate until payment, and costs to be taxed if not agreed.

The third party proceedings

56. The defendant issued third party proceedings against SSEC. The latter, although it gave notice of intention to defend, filed no defence. From correspondence it appears that SSEC will rely on a jurisdiction and arbitration clause in the contract between it and the defendant which provides that the contract is to be interpreted according to the laws of Pennsylvania, U.S.A., and for the reference of any dispute to arbitration in Philadelphia. It will also rely on a legislative provision which gives the Pennsylvania courts a discretion not to enforce a the judgment of a foreign court obtained contrary to an agreement between the parties that the dispute was to be settled otherwise than by proceedings in that court.

57. The question of enforceability is of course a matter for the Pennsylvania courts but does not affect the question before this court. So far as this court is concerned, section 6 of the Arbitration Ordinance Cap 341 provides that a party to proceedings may apply for a stay of the proceedings to arbitration on the ground that the dispute is the subject of an arbitration agreement, after appearance and before delivering any pleadings or taking any step other step in the proceedings. This brings Hong Kong law into line with Article 8 of the UNCITRAL Model Law. No application for stay has been made in these proceedings. There is no provision in Article 8 or Section 6 for stay by the court of its own motion. Therefore there is nothing to prevent this court giving judgment against the third party.

58. Hong Kong law assumes that the laws of all foreign jurisdictions are the same as Hong Law, in the absence of proof by evidence of what the true position is under the foreign law concerned. I have heard no evidence of the difference between the applicable Hong Kong law and the law of the Commonwealth of Pennsylvania.

59. I am satisfied on the evidence of Mr Fok that by an agreement dated 13 December 1995 the defendant was made SSEC's sole distributor of its products in various Asian countries including Hong Kong. SSEC sold the tacker to the defendant, which in turn sold it on to the plaintiff; and SSEC knew that the tacker was to be sold to the plaintiff. The same considerations apply to the sale from SSEC to the defendant, as apply to the sale on to the plaintiff. The tacker did not conform to the description and it was not of merchantable quality. If the defendant sold "junk" to the plaintiff, it had in turn bought "junk" from SSEC, which made it. The defendant is entitled to judgment against SSEC.

60. There will therefore be judgment in favour of the defendant against the third party for an indemnity for the judgment sum, interest and costs awarded to the plaintiff in the main action, and the costs of the third party proceedings to be taxed if not agreed.