Dyson Ltd v Qualtex (UK) Ltd

“Pattern” Parts

Dyson Vacuum Cleaners

The short history of the machines with which this action is concerned is as follows. What follows is a general description of the products in order to provide a context for a consideration of the parts in question in this action. In respect of at least some products the date when it was first made available in the market is important and disputed. I shall make findings about that in due course. In this section of this judgment all dates are approximate. I also include reference to a couple of preceding models in relation to which Mr Dyson had a hand in the design because the prior existence of these models gave rise to certain issues as to the originality of parts of the later disputed designs. Pictures of all the relevant Dyson machines appear in Annexe A to this judgment.

The Iona Fantom and the Cyclon cleaners.

DC01 Upright Cleaner.

DC02 Cylinder Machine

DC03 Upright Machine

DC04 Upright Machine

DC05 Cylinder Machine

DC08 Cylinder Machine

DC11 Cylinder Machine

The Parts in Issue in this Case

PART 2 – THE LEGISLATION

The Legislation – History and the correct approach to interpretation

The statutory provisions and guidance as to their meaning and application

Design right

Design rights in part of an article

Method or principle of construction

Must fit

For the purposes of the sub-section, the article with which the subject article is interfacing can be part of the human body. This might have had relevance in the present case because some of the parts (triggers and a catch) are designed to interface with the human finger or thumb. However, Mr Arnold disclaimed reliance on this – he did not rely on the finger or thumb as a must fit item in the context of any article in the present case.

The sub-section operates to exclude design right even if the relevant part of the design performs some function other than the function described in the sub-section – for example, it is decorative, or has an additional function not falling within the provision. This additional function does not exclude the operation of the provision.

Surface decoration

That distinction may be clear in some cases, but overall it may well be a matter of fact and impression, or a value judgment.

A subsidiary functional purpose does not take the design aspect out of this exemption if the primary purpose is surface decoration. However, the additional purpose has to be subsidiary. This means that I consider that it is likely that this qualification is unlikely to apply in many cases. If an item of decoration has a functional purpose, then it seems to me that it is going to be difficult to say that that functional purpose is sufficiently subsidiary to make the feature surface decoration (or “mere” surface decoration, to use the qualification applied by Park J and Mr Recorder Floyd QC). In truth, I think that the subsidiary purpose of beading that is used to conceal a joint can also fairly be described as decorative.

I can take into account the expressed intention of the designer. Jonathan Parker J obviously did that. This is not determinative, but it is useful material, particularly in relation to functionality.

Commonplace

Design document

Duration of design right

Must match

So far as visual appearance is concerned, Parliament enacted the “must match” provision. In doing so it provided wording which it is not easy to apply. The following elements are relevant and have to be considered.

One has to look at the subject article (“article 1”) and put it in the context of “another article” (“article 2”). The question that arose in the Ford case was whether article 2 was the overall article minus article 1 (viz in that case the whole of the car minus the subject door panel), or whether it was the whole article including article 1 (viz the car including the subject door panel). Both Mr Jeffs QC and McCowan LJ considered the latter approach was correct and rejected the former approach (which they called the “n-1 approach”). While technically those decisions do not bind me on the point (because both were obiter), I shall follow them not least because they seem to me to be right as a matter of logic. If article 1 is intended by the designer to form an “integral part” of article 2, then article 2 must comprise article 1.

Next there is the sense in which the words “integral part” is to be taken. What the provision requires is that article 1 has to be “intended by the designer” to form an integral part of article 2. One therefore has to read in words to the effect of “in design terms” at the end of the paragraph.

That conclusion helps one to tackle the difficult word “dependent”. In what sense is the design of article 1 to depend on the appearance of article 2? It has to be in some design sense – that is the whole context of the provision. But design dependency is a rather fluid and amorphous concept. Most consumer hardware will have been designed externally so as to create what a designer (or the team of designers) sees as a harmonious whole. To that extent any single constituent part might be said to be dependent on the appearance of the whole. In the case of some parts it will be quite obvious to the most uneducated design eye that there are respects in which article 1 is dependent – for example, if there is an obvious curved shape which has to be sustained and bridged across a gap filled by article 1. That, however, will be rare. Absent that sort of consideration there may be a number of designs that would be harmonious, albeit perhaps following a theme. An example of this in the present case is the ribbing on the DC01 wand handle. The detail of the ribbing could be changed (and indeed was changed), without the later design destroying the harmony (dependency) in the earlier, or creating a greater harmony (dependency) in the latter.

Mr Arnold’s approach to this is to say that the dependency is created by the designer. If article 1 is in fact designed in such a way as to make its appearance depend on the appearance of article 2, then the section comes into play. Dependency means no more than that. In particular, it does not import any notions requiring it to be the only way of linking the appearances. This can be illustrated by an example. A car designer creates an overall design, which includes the design of the boot lid. He has design freedom in many of the curves, lines and indentations of the lid (although he will, of course, be constrained by “must fit” matters at the very least). He makes a choice, which, in terms of design, produces an acceptable overall appearance. Mr Arnold would say that the appearance of the boot is “dependent” on the appearance of the whole car, because that is how it appears and that is what the designer intended. Two years later there is facelift for the car. The overall shape is the same, but the designer decides to change the boot a little. Those changes include some changes to the boot – a slight curve change, perhaps, changes in light cluster shapes and perhaps the number plate indent altered in a noticeable fashion. While different from the original boot lid, it still looks as though it “matches” and fits into the overall appearance. Again, according to Mr Arnold, the design of the lid is “dependent” on the appearance of the whole car, because the designer created that dependency when he saw the thing as an integral whole and designed it as such. This example demonstrates two important aspects of his argument. First, dependency does not pre-suppose a single available design solution; and second, dependency is a designer created thing arising out of the designer’s perception and his employer’s implementation of it. Mr Arnold in fact put it round the other way, but it amounts to the same thing. He said the shape of the boot is not independent of the shape of the car, both because it manifestly is not and because the designer did not intend it to be. Independent is the antonym of dependent. Therefore, cancelling out the double negative (“not independent”), the shape of the boot is dependent on the appearance of the whole car.

This argument has some merit. It produces something that is logically coherent, and that is workable. However, it is not wholly consistent with the decision in the Ford case because it would have applied must match to the parts which Mr Jeffs QC thought it was not applicable to. At page 419-420 of that case Mr Jeffs QC seems to have relied on the concept of leaving the appearance of the vehicle unchanged. If a part can be changed without changing that appearance, then there is no dependency and design right can exist in the original part. If it cannot be done without a change of appearance (such as a door panel) then there is dependency within the section. The first of those positions is not consistent with Mr. Arnold’s point; the latter is.

The main problem with this argument, as I see it, is its logical end and the paradox which it seems to me to introduce. As I have already observed, most modern consumer goods are likely to have been produced from a designer’s pen (or CAD package). If made up of more than one part (which, again, most will have been) then each of those parts is likely to have been designed with the others in mind, and to fit in. In Mr Arnold’s reading of the section, the design of each will be dependent on the appearance of the whole because that is how the designer will have intended it. The result of this is that the external aspects of most external parts will be exempted from design right. That would make very large inroads into the rights that the statute creates, and might even be said fundamentally to undermine them. That might be said to be justifiable and sensible if one’s point of view is that the consumer must be free to replace or repair those parts without being confined to the manufacturer’s parts, but I have already indicated that if that was Parliament’s intention it could have chosen some easier words to achieve it. Furthermore, it leads to a paradox – the more thoughtful the design, the less likely it is to attract design right. Design rights are presumably intended to give protection to the fruits of the design process. A good design process will be designed to combine separate parts into a harmonious whole. The more effort that is put into that, the more successful (and probably the more expensive) the exercise will be. Yet the more successful it is, the more likely it is that there will be the dependency that Mr Arnold relies on. Thus separately designed parts thrown together in a casual non-thought-out fashion will each attract design right, while the carefully combined parts of the conscientious aesthetic designer will not. Mr Arnold did not really accept that this paradox existed, or at least not as starkly as I have presented it, but I consider that it does. Since it seems fundamentally to undermine the very right that the statute seeks to create, it causes me to question Mr Arnold’s analysis.

Accordingly, when considering “must match”, I shall apply section 213(3)(b)(ii) by considering whether there is dependency of the kind, or to the extent, which would make the overall article in question (article 2) radically different in appearance if article 1 were not the shape it is. That is not to substitute, or superimpose, a test different from that appearing in the statute. It is to explain how the statute works. This sort of consideration is inevitably one of fact and degree, and of impression. It will doubtless be less than straightforward to apply in relation to various parts or aspects, but that is the nature of the beast. The saleability of the item can be used as a sort of guide to assessment or a cross-check in any particular case.

PART 3 – WITNESSES AND GENERAL EVIDENCE

Witnesses

I also heard from three expert witnesses – Mr Jupp (who gave expert evidence on design in relation to all parts), Mr Hulme (who gave expert evidence on design in relation to all parts except brushes) and Mr Hadfield (who gave expert evidence on brush manufacture and design). Neither of the first two of them was fully independent in that they both had or have significant relationships with the parties (as a result of which both gave factual evidence as well). Mr Jupp was Dyson’s expert. He holds a manufacturing engineering degree and for almost 4 years between 1988 and 1992 was employed by Dyson. In 1998 he resumed full time employment within the Dyson group and since February 2001 he has been Group Product Director of Dyson Limited. In the intervening period he was a freelance design consultant, and for periods worked for Dyson (including a period working on one of the smaller parts which are in dispute in this action). He is thus closely associated with Dyson. Despite that, I found him to have achieved and demonstrated a genuine independence of view and to have discharged his responsibilities as an expert witness well. He expressed himself moderately and carefully and in no way indicated that he was an advocate in a cause as opposed to an independent expert doing his best to assist the court. He readily made concessions where concessions were required, and where he had not thought of certain things, and did not consider himself to be an advocate in his employer’s cause. I was greatly assisted by his evidence.

Mr Hulme has spent his adult life in the electrical repair trade. He started out repairing things himself, and moved up through his various employers’ companies. In 1961 he started the defendant company, and over the years he has built it up so that it is now a very substantial company. He is its chairman and principal shareholder. He estimates that it has almost half of the vacuum cleaner pattern spares market in the UK, and it operates the second largest pattern parts warehouse in Europe. It is clearly a substantial company and a big player in the pattern parts market. Mr Hulme is also a significant figure in that market with strong views on the need to have a pattern parts market and the undesirability (as he sees it) of letting OEMs have a monopoly, which he regards as having elements of bullying and oppression. On two or three occasions in recent years he has assisted pattern parts suppliers in their disputes with larger OEMs. He has obviously done so because he believes in the cause. He sees the consumer as having to pay unnecessary millions of pounds per year more if the OEMs’ claims to design right (and other rights) are not kept in check. His battles took him into court against Hoover in 1982 (Hoover plc v George Hulme (Stockport) Ltd [1982] FSR 565). In that case, and significantly for the present case, he succeeded on a waiver defence.

There is nothing intrinsically wrong in Mr Hulme’s hositility to OEM monopoly in the spare parts trade, and he has impressed me as an entirely honest and honourable man. However, I have obtained the firm impression that his attitudes have been allowed to colour his objectivity in this case so far as his role as an expert is concerned. His views do not necessarily make him an inappropriate expert witness, but they risk imperilling his objectivity, and in my view that has, to an extent, happened in this case. I think that from time to time he tended to don the mantle of a crusader or advocate, rather than that of an expert witness, and there are times when I think that his evidence of opinion (particularly on commonplace) has to be treated with considerable caution.

Both Mr Jupp and Mr Hulme gave some evidence of fact as well as their expert evidence. I accept Mr Jupp as an honest and truthful witness of fact. Mr Hulme was also honest in his factual evidence, though his hostility to the interests of the OEMs sometimes overlaid a certain gloss on his evidence.

The other expert was Mr Ronald Hadfield. He is the technical consultant to the Accrington Brush Company Ltd (“ABC”) and has a considerable background in engineering. He joined ABC as an engineer in 1969 and after 18 months he was promoted to general manager. When he reached retirement age (65) he was asked to stay on to help develop younger employees. There can be no doubting his expertise in technical brush-making matters. It was to ABC that Dyson turned for bristling matters (and a certain amount of input at the design stage), and Mr Hadfield gave some factual evidence as well as expert evidence on bristling matters. In addition to his expertise being beyond doubt, his honesty was beyond doubt as well and he was brutally frank about some of Dyson’s bristling ideas. He was a helpful and credible witness, and so far as I do not accept the full force of everything that he said to me (which is rare) it is only because of what I think are differences of perception rather than unreliability or an attempt to mislead.

Dyson machines as design icons

PART 4 – DESIGN RIGHT APPLIED TO THE PARTS

The subject parts – generally

From time to time Mr Carr complained that Qualtex’s case was what he described as a moveable feast. The Defence made certain claims as to the applicability of the design right exceptions to various parts. Mr Hulme’s various reports often went further in some respects and sought to apply them to additional features. Mr Jupp’s report met the pleaded case. Mr Carr was happy to take Mr Hulme’s report as being the final position on which he was entitled to rely, and not take a pleading point in relation to differences. However that was where the line should be drawn, he said, and Qualtex should not be allowed to go beyond that. On occasions Qualtex’s case sought to do so. I accept that as a general approach, and will myself adopt it, though with a certain measure of flexibility where I think I can fairly deal with the point anyway.

1992 DC01 wand handle

The details of the relevant parts of the wand handle can be seen on the photograph appearing in the Annexe. I need to describe the detail.

Starting at the top, its construction is as follows. At the top there is a partially ball-shaped bulge, with ribbing on it. In the two middle valleys between the ribbing there are two slots in the front surface, and two in the back, which connect the interior of the bulge with the exterior.

If one looks down at the top of the bulge, one can see the upper end of the tube which defines the airpath up from the lower areas of the handle. There is a small clearance between the outer wall of this tube and the inner wall of the bulge. The walls of this inner tube are not complete cylinders. At the top of the tube there are two opposed semi-circular cut-outs which allow the air at the end of the tube to communicate with the air in the small clearance area around the inside of the bulge. This, combined with the slots referred to in paragraph (ii), is a safety feature. If the end of the handle becomes blocked, air can still pass down the tube via the slots and the cut-outs – it is a bleed feature. This prevents the motor burning out, and also guards against some of the dangers to an eyeball if it were pressed against the end of the tube (a gory event which the designers apparently had in mind but which fortunately they had never heard of as occurring). When a tool is fitted into the end of the wand handle its rounded connecting tube fits over the inner tube shape, so that it fills the gap between that tube and the inner wall of the bulge.

Below this bulge and at the rear of the handle is a series of grooves cut into the surface which would naturally be gripped by the user’s hand. These are said to be to enhance the grip.

Below that there is a series of ribs. At least one of the purposes of these ribs is to act as a design feature, to disguise the change in width of the handle (which is apparent from the drawing). There is a dispute as to whether they are there to enhance grip.

Below those ribs, at the rear of the handle, is a short horizontal tube feature. This tube is to form something around which the mains cable can be wound (it is wound round a tab towards the bottom of the machine at the other end) and to seat a rotating tab whose “legs” sit in the tube and which, when rotated vertically, stops the cable slipping off. The whole feature is a cable winder; the tubular feature is the cable winder boss.

Opposite that feature, on the front, is a release catch. The catch is hinged to the wand handle at the handle’s lower end. It sits around, but not touching, the lower part of the handle, and is permitted to move towards the handle when “squeezed” by the user. When it does so, a lip on the front of the catch disengages from the body of the machine, allowing the release of the wand.

Each of the catch and the handle has a four-legged, cross-shaped (in plan view) opposing spigot. A spring sits over both spigots, resisting the movement of the catch when the machine is released and springing it back when the movement is finished.

Beneath the area of the spring, and around the spigot, is some additional ribbing. This ribbing was not produced on later versions, and is not material to this action. I need not deal with it further.

Mr Arnold submitted that apart from the bulge at the top (though the ribbing there is said to be non-original), and the cable winder boss (which does not appear at all on the Cyclon handle), the 1992 wand handle is not original over the Cyclon handle. He submitted that once the derivation is admitted (which it was), the next question was whether any differences were visually significant. If they were not, then the later design was not original. Differences have to be visually significant for there to be originality. I have already decided against this argument in the context of considering originality above. It therefore does not assist him. Furthermore, even if I were wrong about that, then he would fail on the facts. The Cyclon wand handle was exhibited in these proceedings, and I have had a chance to consider it closely and compare it with the 1992 DC01 wand handle. I am quite satisfied that the differences between the two are visually significant. So even if that were a qualification for originality in this case, the 1992 handle design possesses it.

There was originally a suggestion that parts of the 1992 handle were derived from other predecessor machines (principally the Iona Fantom). However, that source of derivation did not really figure by the time of final speeches. So far as there was derivation, it was derivation only and the attack on originality fails for the same reasons as it fails in relation to originality over the Cyclon handle.

1992 DC01 wand handle – must match

To my mind and eye the must-match provision does not apply to this part. Changing one or more or all of the design features of this handle would not necessarily make the whole machine (including the handle) look “radically different” within the guidance that appears from the cases. I do not think that this part is like the door panels in Ford. It is of course true that one could make the handle so different that the overall appearance of the machine would be radically different, but that is not necessarily the case. One could alter the detail of the bulge at the top, or not have a bulge at all, or have different ribbing details lower down the handle, or have different (or no) grooves on the underside where the hand naturally rests, without radically changing the overall appearance of the overall machine within the meaning of that phrase. But it does not follow that any change would make the handle radically different. So far as one can make a cross-check by inquiring what sort of a market there is or would be for a non-replica handle, while it is true that there was no evidence of any such market, there is no evidence that there would not be one. In the absence of solid and reasoned evidence (which in the circumstances would have had to have been opinion evidence) I do not think it is possible to say that no-one would buy a non-replica handle, and I would be a little sceptical of assertions to that effect. It is conceivable they might, if there were a price differential which made it worthwhile. As I have said, I do not think the handle is like a car body part. Cars have a certain public display element to them. They are seen in public, and customer choice is likely to be affected, at least in part, by the design and the customer’s willingness to be associated with the design by driving in it or having it parked in his driveway. The same cannot be said of vacuum cleaners. A Dyson vacuum cleaner may well be purchased at least in part because of its design, but I would require some evidence were it to be said that the design preference of the customer for this piece of household utilitarianism would lead the customer to require it to keep its looks after a repair in the same way as a car is required to keep its looks. This is, I accept, somewhat speculative in the absence of solid evidence, but it is plausible and I certainly cannot find that that is wrong and Qualtex has not discharged its burden of showing it is right.

Qualtex’s secondary case on must match takes certain specific features and alleges that they fall within the exemption. They are as follows:

The next aspect of must match relied on in relation to this handle is what is described as a half lozenge (curved) shape at the back of the lower part of the handle. Mr Hulme said that this was designed to match the equivalent shape on the front of the machine so as to form a complete lozenge shape if viewed from above and in cross-section. Mr Jones accepted that there could be said to be a match, and Mr Jupp agreed there was a “congruity” but denied suggestions that there was a match. It seems to me that Mr Hulme’s evidence, and the cross-examination of the two Dyson witnesses just referred to, misses the point. The point is not whether there is a match. The point is whether there is a dependence for the purposes of the statutory provision. In my view there clearly is no such dependence. There is no reason why that particular aspect of that particular part has to be the shape that it is because that shape is dependent on the appearance of the whole machine. It would not radically change the appearance of the machine were this part of the handle to be another shape.

Last, Qualtex rely on the shape of the rotatable “hook” on the cable winder. Mr Hulme says that this matches the round shape of the boss into which it fits. That is true; but its shape is not dependent on any aspect of the design of the overall machine in the sense required by the section. This point fails.

1992 DC01 wand handle – must fit

The semi-circular or curved part of the handle which lies underneath the handle release catch. Underneath the release catch there is a wall of the handle which has a curve parallel to the underside of the catch. They do not touch, even when the catch is pressed – there is a small clearance. As will be apparent later on, Mr Jupp accepted that the curved under-surface of the release catch which lies above and roughly parallel to this surface of the handle was a must fit feature of the catch. However, he resisted the notion that logically this made the surface of the handle must fit as well. He pointed out that the surface of the handle did not have to be that shape at all. I do not think that it is necessarily relevant whether the underlying wall had to be that shape. The fact is that it is that shape. However, I think that there is a different reason why the catch is must fit and the handle is not. Looking at the matter realistically I think that while it can be said that the catch is the shape it is to enable the catch to perform its function, I do not think that the evidence allows me to say that the handle is the way it is to enable the catch to perform its function. Accordingly this aspect of the handle is not must fit. The design history of the parts showed that the handle was designed first and the catch was designed later. The handle had its first expression in a design drawing in February 1992 and the catch found its first expression in a drawing in April 1992. While those are formal drawings which do not necessary reflect the precise form of the order of designs, in my view they are likely to do so, thus making the catch must fit but this part of the body of the handle not must fit.

Qualtex put a second must fit case in relation to this aspect of the handle. When the catch is at rest there is a space between the underside of the catch and the curved surface in question – they are approximately parallel when the catch is at rest. It was said that the bottom part of this wall abuts the catch when a user holds the wand by the catch for above the surface cleaning, and that it does this so as to stop the catch flexing and “popping off” its pivot pins. I do not consider that this qualifies. Mr Jupp did not consider that this was a design consideration, but in any event did not accept that this was the effect of that bottom part of this surface. He described the suggestion of “popping off” as conjecture, and I agree with him that in the absence of experimentation (of which there was none) it was. In those circumstances I do not consider that this is an aspect of the shape or configuration of the handle which falls within any aspect of the section; and the same is likely to be true even if the “popping off” theory were established – in that event there would only be a limited contact and the whole feature still does not fall within the section.

The clearance around the spring to stop it fouling. This refers to the area of space around the spring against which the catch moves; the area enables the compressed spring to expand into space and to deflect slightly to the side (it deflects slightly because the catch describes an arc rather than having an entirely straight line movement, because it is pivoted at one end). It was claimed in Qualtex’s final speech but was not the subject of any pleaded claim. Nor did Mr Hulme give any evidence in relation to this, and there was no cross-examination on it. I therefore make no ruling on it.

A small area between the stops which prevent outward movement of the handle release catch and the top face of the catch. This requires explanation. The catch is sprung so that, unless restrained, it would be pushed away from and off the handle. It has to be restrained. That restraint is provided by a short vertical protruberance at its top which, when the handle is at rest, presses against a matching short vertical downward protruberance at the bottom of the ribs of the handle. When they meet the handle must stop moving. There is a clearance between the latter protuberance and the top face of the catch. This is the clearance referred to. Like the clearance around the spring, it is not an aspect that formed part of Qualtex’s pleaded case, and it did not form part of Mr Hulme’s evidential case either, though it was referred to in the cross-examination of Mr Jupp. In the absence of a pleaded or evidential case in respect of this clearance I am not minded to allow Qualtex to rely on the point, but in any event such evidence as Mr Jupp gave on it did not support the case relied on in final submissions and as a result I do not think it qualifies. The cross-examination did not go so far as to establish that this clearance was a particular aspect of the design capable of falling within the section. Accordingly I rule against Qualtex on this particular aspect. However, I suspect that in any event this alleged point is not material since it is accepted by Mr Jupp and Dyson that the stop itself is within the must fit exemption, and to some extent the resulting clearance is an aspect of the stop itself.

The forward facing surfaces of the vertical ridges which interact with the vertical splines on the machines, which must be clear of the handle release catch. The body of the machines contain a vertical spline. When the handle is slotted into the body this spline locates itself into a groove formed by the curved back of the handle as it comes round to the front and by a vertical feature (called a ridge in cross-examination of Mr Jupp – it is like a single raised rib standing proud of the surface of the handle). The rearward facing face of this ridge forms one side of the groove. The forward facing surface does not engage with anything. Qualtex say that this surface falls within the must fit provision because it has to be a sufficient distance from the handle release catch to enable the catch to travel the required distance when depressed. This is said to mean that this surface is a feature or aspect which enables the catch to perform its function. I do not consider this submission to be correct because I do not consider this to be the correct analysis. The surface itself is where it is and what it is. It is merely the reverse side of the ridge whose function is to form a side of the groove. It is true that if it were somewhere else (farther towards the catch) it would foul the catch, but this does not mean that the surface is a feature which enables the catch to function, or which enables the ridge to form the groove.

Flat forward-facing surfaces above and behind the locating ridges at the lower end of the handle. The argument of Qualtex on this is the same as in the preceding paragraph – these surfaces must be clear of the release catch when depressed. I reject this argument for the same reason, though the position is even clearer in relation to this “aspect” of the design. This surface is the finishing point of the rear curved section of the handle. It is true that if it came too far forward it would foul the release catch when the latter were depressed, but that does not make this surface a feature which enables the handle to be connected to or placed around or against the catch so that the catch may perform its function. It is simply the way the designer chose to finish the curved surface. It cannot sensibly be treated as an aspect of the design which enables either part to perform its function.

Next are features which enable the handle to fit into the machine. The contested elements are as follows.

The vertical grooves which receive vertical splines on the machine. I have referred to these already. Dyson does not dispute that part of this groove is within the exception but points out that the defined groove is longer than it needs to be by a couple of centimetres or so. This length is visible on the handle when it is engaged. Where one has such a feature I think that one has to apply the Act realistically and pragmatically. If the groove were, say, 2 mm longer than the spline then it would in reality still be the same groove and be within the must fit exception. At some point one can no longer say that – where that point is is a matter of fact and degree. I think that on the facts of this case the upper part of the groove is still part of the same groove and within the exception. Qualtex succeeds on this point.

The flat bottom of the handle which is said to sit on a horizontal surface at the bottom of the socket on the machine which receives the handle. It is curved to match the curve of the handle (actually the catch, at this point). At the front of the handle there is a curved section which is covered by the handle release catch. That runs down to the bottom of the handle. When the handle is placed in the receiving socket on the machine something has to stop its descent. This is achieved by having a horizontal surface (a sort of shelf) on the machine which provides the stop – there is actually a raised nub on it. At the frontmost part of the curve of the handle (underneath the catch) there is a raised nub - it is about 4 mm square and about 1 mm in height. When the handle is placed in the socket that nub contacts a nub on the shelf and together they stop the descent of the handle. The combination of those features means that the bottom of the handle’s curved section does not actually sit on the shelf. It is the two nubs which engage. Nevertheless Mr Arnold says that I should adopt a “practical approach” and say that the whole of the bottom of the handle falls within the must fit exception, and not merely the nub. The nub is clearly a feature which enables the handle to be placed in the machine so that both can perform their respective functions. It is obviously a deliberately placed item – it is apparent that there is intended to be no touching between the shelf and the bottom part of the handle other than at the point of the two nubs. I do not think it is possible to say that any other part of the bottom surface enables the handle to be connected to or placed in, around or against the body of the machine so that the handle (or the machine) can perform its function. The appeal to the “practical approach” is an appeal to ignore the words of the section and the design realities of the part. The section, and the approach of the parties in this case, has required a breakdown of parts into their various aspects, and one cannot move from one aspect which complies to another which does not merely on the grounds of practicality. The point was not put to me on the footing that the curved section of the handle above the nub was somehow the male part of a male/female joint, and on the facts I very much doubt if that could be done. Short of such an argument, I cannot find that the bottom face of the curved section (which is the part relied on by Qualtex) falls within the must fit exception.

The interior of the tubular boss which receives the cable winder. Although not clearly accepted on the pleadings, Mr Jupp accepted in evidence that the inner surface of this boss, and the internal pocket into which the winder legs clicked, were within this exception and I so find.

The upper half of the tubular boss and the immediately vertically adjacent profile of the ribbing. These are said to be must fit because the cable is wound round the boss and is restrained from slipping inwards by the adjacent face of the ribbing. It is therefore said that these are features of the main wand handle which are placed against the cable enabling the boss to perform its function (of receiving the cable when wound round it). There was no material dispute about the surface of the boss, though Mr Jupp did observe that the exact parts of the boss surface which interfaced with the cable was debatable and not necessarily the full semi-circle claimed by Mr Hulme. However, I accept Mr Hulme’s evidence and find that the top half of this boss is within this exception. Although there was little debate about it, I find the same about the vertical surface of the ribbing immediately adjacent to (and above) this boss. It is this which restrains lateral movement of the cable; it performs the same function on that side as the tab (to which I will come in due course) performs on the other, and is within this exception.

Features in the bulge at the upper end

1992 DC01 wand handle - surface decoration

The lower ribbing area. The first area of the ribbing is the pronounced and obvious area starting just above the cable winder boss and running down to the handle release catch. The evidence of Mr Jupp was that this was there to hide the change in diameter of the two main parts of the wand handle that occurs in that general area. Mr Jones, the designer, said that their purpose was to create something of an optical illusion – to make the handle appear more substantial there than it in fact was. There is a slanted rib running at right angles to the horizontal ribs at the front of the handle and parallel to their top surface. This rib was, according to Mr Jones, to give the impression of a stronger handle. Mr Arnold submitted that all these features demonstrated that the ribbing was surface decoration (though the vertical rib was not pleaded, nor did it appear in Mr Hulme’s evidence). I disagree. I think they demonstrate precisely the opposite. The ribbing is carefully designed to convey a shape, defined by that ribbing; it does not merely decorate a surface which is already there. It goes far beyond that. Even Mr Hulme accepted that without the ribbing the shape of the handle would be completely different. I find that this ribbing is not surface decoration. Mr Jupp suggested in his expert’s report that in addition the ribs provided an extra gripping area when tools were being attached or removed from the end socket. I doubt if that is the case – it is not a natural place to hold the handle when performing that activity. In his final speech Mr Carr did not rely on this alleged functional feature of these ribs. In the circumstances no additional functionality point arises in this context.

The ribbing on the bulge at the end. Mr Jones, who designed it, told me that the ribbing was to assist grip. There were two situations in which this might be necessary or desirable – first when manoeuvring the machine around the floor in some circumstances (using it as the ball to the hand’s socket), and second when removing a tool. Mr Jupp agreed the ribs promoted grip, and also said that the ribs made manufacture easier (by reducing the possibility of sinking or distortion). He pointed out that they provided a pathway for the bleed holes in the grooves between the ribs. Mr Hulme, in his report, merely asserted that these ribs were surface decoration. In cross-examination he disputed the grip point, did not dispute (but did not accept) the manufacturing point and accepted the shape would be different without the ribs. He agreed that they provided a pathway for the bleed holes. Taking into account this evidence, and my own impressions, I find that these ribs are not surface decoration. They fulfil practical functions (grip and the airway passage) and define the shape of the bulge, and those functions are in no way subsidiary to some other function. They are not decorating a bulge that is already there as part of the design.

The grooves on the lower general gripping part. The handle is obviously designed so that most of the time the usual way of gripping the handle is by the tubular section below the bulge (without in any way derogating from my finding that on occasions the bulge will be used). At the back of this section there are 10 grooves, decreasing in depth on each side as they move towards the front. Qualtex says these are both decorative and functional, and that this makes them surface decoration. The evidence of Mr Jones was to the effect that these grooves were to facilitate gripping. In order to give them a more pleasing effect he tapered them off, but that did not detract from their function of facilitating gripping. Mr Hulme stated that he did not accept that they facilitated gripping to any significant extent, and thought that they were decorative. Part of his reasoning was that Dyson dropped the groove on later models. That is true in a sense – the DC04 and DC07 do not have these grooves, but they have a very different handle configuration.

I have no difficulty in finding, as I do, that the designer intended these grooves to provide grip, though he obviously intended to do so in an aesthetically pleasing manner – hence the tapering or chamfering. I also find that they do provide some additional grip in fact. It may be that it is the case that they also advertise to the user that this is where the handle is to be held (though the user would doubtless come to the same conclusion in their absence) but that is subsidiary. In the light of those findings I find that these grooves are functional. They are therefore not surface decoration. Even if they had a decorative aspect I would find that this aspect is very much subsidiary to their functional aspect, so that they are still not surface decoration, in line with the principles I have extracted above.

DC01 1992 wand handle – commonplace

However, Mr Hulme also applied a commonplace test to various of the elements of the handle – grooves, bulges and so on. In doing so he produced evidence of various examples of those features found on historical machines, designs and patents, and used them as evidence that the particular aspect was itself commonplace. In the main he has gone about an exercise which does not help. To take an example, in relation to the bulge on the wand handle, he has produced examples of bulges in a photograph of a very old vacuum cleaner which bears no resemblance to a Dyson machine at all, the bulge apparently being on a manual pumping handle (so far as I can tell). Then he produces a line drawing, taken from a US patent, of a vacuum cleaner hose with some sort of bulge in the middle of it (like the bulge one would see part way down a cartoon boa constrictor a short while after a recent moderate meal); and then he produces a diagram for the old Dyson-designed Iona Fantom cleaner which shows a bulge at the end of a wand handle. That material shows bulges. It does not show them to be commonplace in vacuum cleaner design. The first two examples were documents taken from the internet. There was no evidence as to how widespread the use of the relevant machines was, or even whether they were ever in use at all, so even assuming the bulges could be related to the bulge at the end of the DC01 wand handle there was no material linking them to designer-awareness so as to assist in establishing commonplace in relation to the bulge. Mr Hulme frequently did this. He confused evidence that someone had once thought of a design concept in one context with the evidence of general designer-awareness (whether through marketing or otherwise) on the other. Only the latter could be relevant. Mr Hulme and Qualtex have failed to establish it in relation to any of the aspects of the DC01 wand handle which were said to have been commonplace. Thus:

It was asserted that grooves on a gripping handle were commonly used before the Cyclon, and were used on the Cyclon. A line drawing of a Hoover Junior 1334 machine, said to have been very common in the UK, Australia and (in an early version) the USA, was produced, apparently showing an area of grooving on the handle. The grooves, so far as discernible, are much closer together than the DC01’s and the handle is different so they are not in a clearly analogous position. Mr Hulme said they were decorative only; they did not enhance the grip. This is not sufficient evidence to show that grooves like those on the gripping part of the DC01 wand handle were commonplace in the sense of the statute. The Hoover grooves were clearly different, and the presence of the grooves on another Dyson machine does not make them commonplace or add any evidence demonstrating that they were. Mr Hulme also asserted that there were grooves on other models, but his evidence on this was not sufficient to demonstrate commonplace.

I have already referred to Mr Hulme’s evidence of bulges. He did not assist his case when he said in his report: “Although bulges on handles were not very widespread …”. He has produced no evidence which would go anywhere near supporting an assertion that bulges like that on the wand handle were commonplace. All he has produced is evidence of different bulges on different parts of different machines which may or may not have had some commercial success.

He said that ribbing had long been used to adorn generally standard vacuum cleaners and he produced pictures of an Electrolux stair tool which showed a small amount of ribbing on top. That, however, is insufficient. Showing that ribbing was commonplace on vacuum cleaners does not show that the aspects of the Dyson design which featured ribbing were commonplace. The ribbing on the DC01 wand handle cannot for these purposes be properly described as “ribbing on a vacuum cleaner”. It is particular ribbing with particular dimensions and particular aspects. That is the design (or designs) over which design right is claimed. Ribbing like that is not (on the evidence) commonplace. That conclusion is not particularly surprising.

1992 DC01 wand handle – method or principle of construction

1992 DC01 wand handle - generally

Infringement

DC01 wand handle release catch

DC01 wand handle release catch - must fit

The surface at the top across which the stop on the handle moves. This is an area which was not pleaded and which is said not to have formed part of the evidence of Mr Hulme. This is largely true. While Mr Hulme’s drawings seem to colour this surface as must fit his report does not refer to it and there was no cross-examination on the point. I therefore decline formally to rule on it. However, I would add that my impression is that it would not qualify as a must fit item. This surface does not enable the catch to be placed in, around or against the remainder of the handle. The point made by Mr Arnold in his final submissions is in effect that it must not foul the stop. However, this surface does not “enable” non-fouling. It is simply out of the way and no more enables the catch and handle to operate properly together than any other part of the handle or catch (or indeed any other more remote part of the machine). It would be a misuse of the language of the section to say that this surface fell within it.

The edges of the curved section at the bottom. Mr Jupp accepted that both inner and outer faces of the curved bottom of this catch were excluded, but not the shape of the narrow edges. I think that this an artificial distinction. If the two surfaces are excluded, then the edge is in essence a function of the thickness of the plastic. If the main surfaces are must fit, then in reality the edges must be as well. I find against Dyson on this.

The curved edges of the portion of the catch around the pivot holes. These are the surfaces which define the edge of the catch in the area of the pivot holes. They are said to be the shape they are in order to clear two things. The first is the splines on the body of the machine which engage with the groove higher up the handle, as observed above. The splines pass immediately by those edges before they engage in the groove formed by a plastic ridge on one side and the edge of the curved back of the handle on the other. In practice they form, or almost form, the beginning of one side of the groove. The shape of this part of the catch has to be clear of the splines in its position at rest and when rotated slightly by the depression of the catch. The second feature is the bottom edge of the ridge itself – it has a curved shape which matches the curved shape of this part of the catch, though they do not quite engage.

Mr Jones accepted that this piece of local detail was dictated by the splines on the body of the machine, and also by the need to clear the ridge. Mr Hulme pointed out that these edges needed to be clear of the splines when depressed, but did not otherwise describe any need to interface. Mr Jupp was cross-examined more on clearances than on enablement within must fit.

On this point I find in favour of Qualtex. The plastic around the pivot hole obviously has to have more than an ephemeral width. The shape of the plastic at this point is dicated by the need to allow passage of the spline and rotation of the catch; and it has at some point to fit in with the vertical ridge. These findings mean that this aspect of the design of the catch enables it to be connected to or placed against the wand handle in such a manner as to allow it to perform its function (to rotate around the pivot pins when pressed) and to allow the handle to perform its function of fitting into the body of the machine. It also allows the splines on the body of the machine to pass immediately by it and engage in the groove on the wand handle, thus enabling these splines to perform their function. I therefore find that this aspect is caught by the must fit exception.

The flat bottom of the catch that fits into the horizontal surface of the machine body immediately beneath it. This is the same point as arises in relation to the bottom surface of the curved section of the wand handle, which I have dealt with above. I have described above how the handle is stopped by a nub that meets a corresponding nub on the “shelf” in the body of the machine. The catch surface is essentially parallel to the curve on the handle, and does not extend below it. It follows that this bottom surface does not engage with the “shelf” at all – it is deliberately kept clear of it. Accordingly, it follows that if one looks at this surface as a separate aspect for these purposes it does not qualify as a must fit aspect for the same reasons as the corresponding surface on the wand handle. However, it seems to me (although it was not argued this way) that must fit can be put another way in relation to this surface. It is accepted by Dyson that the whole of the lower part of the release catch (both inner and outer surfaces) are within the exception, the former because it fits into the matching aperture on the body of the machine and the latter because it is positioned around the main wand handle when the handle is depressed. The bottom surface is in effect an inevitable part that links the front and back surface. It is not sensible to treat it as a separate aspect for these purposes. If the front surface is within the exception, then its natural termination must be (the termination must have some recognisable thickness); and the same applies to the rear surface. The end surface is the natural termination of both. Accordingly, it falls within the must fit exception as much as both surfaces do.

DC01 wand handle release catch - must match specific features

It is true that there are matches. The profile, pitch, spacing and angle of the ribs is the same. In its “at rest” position the handle stands proud of the line of the top edge of the wand handle ribs, but when it is depressed the catch ribs come almost into line with the handle’s – if one presses hard they go beyond alignment to a slight non-alignment again. If matching were a test, the catch would fulfil it. However, it is not the test, as I have pointed out above in relation to similar arguments in relation to the handle. Everything I have said there can be applied to this ribbing. I do not think that replacing this catch with a catch of another design would inevitably make the handle (or machine) look radically different. I therefore reject Qualtex’s argument and case on this ribbing.

DC01 wand handle release catch - surface decoration

DC01 wand handle release catch - originality

DC01 wand handle release catch - commonplace

DC01 wand handle release catch - infringement

DC01 1992 Cable Winder

No design right is claimed in this piece of the overall handle as a separate item. This is probably because the pleading (if not Mr Jupp’s evidence) was that it was copied from the Iona Fantom cleaner. However, in the light of the claim to originality in the whole handle including the cable winder it is necessary for me to consider other attacks on design right based on must fit and must match.

DC01 1992 cable winder - must fit

Qualtex claimed that the rear surface of the cable winder, which held the cable in place, fell within the must fit exception. I find that this part does indeed fall within the must fit exception. It enables it to be placed against another article (the cable) so that it can perform its function.

The semi-circular shape at that part which abuts the boss also falls within the must fit exception. When the winder is rotated downwards, to permit the cable to slide off, this shape lines up with the shape of the upper half of the boss. If it protruded beyond the surface of the boss it would snag the cable and the cable would not be able to slide off so cleanly. I find that this feature of its shape enables the winder to perform its function of releasing the cable when required to do so.

Beyond that there are no features to which the must fit exception applies. Qualtex sought to argue that the sides of the winder fell within the exception because if they were wider they would foul the cable when it was slipped off. Mr Jupp conceded that this was a possibility if it were “much, much fatter”, but I do not consider that that means that this particular design of the winder falls within the exception. The fact that, if designed differently, it would not work is not sufficient. The shape is not one which “enables” the winder to perform its function. The most can be said is that it does not obstruct it, and that is not enough, in my view.

DC01 1992 cable winder - must match

DC01 1992 cable winder - commonplace

DC01 wand handle - the whole handle (including the catch and cable winder)

I reject this submission of Qualtex. The fact that the cable winder was copied from the Fantom means that that article is not original for the purposes of design right. I do not think that that deprives the combination of originality. In Ultraframe counsel for the defendant is recorded (at para 71) as not disputing that a combination of features that are commonplace could still be the subject of design right, and Laddie J did not question that concession. I would respectfully agree. Commonplace goes to originality – see section 213(4) – and in my view the same goes for items which are not original because they are copied; that is to say, design right can exist in a combination of features some of which (or even all of which) are copies when taken separately. Reverting to the present case, therefore, the fact that the cable winder is a copy does not prevent design right from arising in the whole of the handle assembly of which it forms part, provided the overall design is original (which, in my view, it is).

However, in considering the extent to which design right exists in the whole assembly, those parts which fall within the must fit and must match exceptions have still to be excluded. This was conceded by Dyson (who, on the true construction of their pleadings, did not claim rights in anything more than the combination minus must fit features), and it is clearly the approach of section 213(3)(b) which clearly applies to “features of shape or configuration of [the] article”, the “article” for these purposes being the overall handle. It was also the approach adopted in the pleadings without criticism in Farmer’s Build (see p 469).

I find, therefore, that design right is capable of existing in the whole handle assembly save for those features which are conceded as being, or which I have found as being, within the must fit exception (there being none which fall within the must match exception). In the light of my other findings I find that right to exist.

DC01 wand handle - infringement

DC01 1995 Wand Handle design

Dyson’s position in relation to this handle is as follows. Much of the handle design is the same as the 1992 version. So far as that is the case, Dyson does not claim a new design right arising in 1995. It relies on its design right arising in 1992. However, so far as the design has changed, then it relies on new design right in the changed parts (so far as not excluded by any of the provisions of the Act). The extent to which design right exists in relation to the 1992 aspects has already been dealt with by me in considering the 1992 handle. That means that only the new internal arrangement, the different ribbing and the changed position of the cable winder boss have to be considered. In relation to those aspects the following is the position:

So far as must match is concerned, although the ribbing detail is different, the legal argument, and the legal result, is the same as in relation to the ribbing on the 1992 handle. It is not excluded on the footing of must match for the same reasons.

Where ribbing has been removed on the 1995 version, the “lozenge” shape of the back of the machine has replaced it. This, said Qualtex, makes this aspect of the handle non-original. I reject that submission. It was not copied in the “tracing” sense, and so is original.

The DC01 1997 Wand Handle design

Dyson claim design right in relation to the tab only. Dyson and its expert Mr Jupp between them accept (and so far as necessary I find) that the exterior surface of the winder tube (which fits into the boss), the receiving part of the boss and the interior surface of the recess which receives the screw are all covered by the must fit exception. That leaves the tab, which has to be covered separately. Qualtex asserts that design right does not exist in the tab for a number of reasons, which will by now have become familiar.

Originality. This time there is no admission of copying of any form. In fact, the widening of the tab was not mentioned by the designer, Mr Thomson, who gave evidence. The point emerged in his cross-examination. It was not put to him that he had merely copied the 1992/1995 tab, and he did not say that he had. In the circumstances there is no direct evidence of copying in the relevant sense. I am not prepared to draw an inference of copying. Accordingly, this time the tab design is original for the purposes of unregistered design right.

Must fit. I have already recorded the concessions and findings in relation to the tube that fits into the boss. In addition the same aspects in respect of which must fit was successfully claimed in relation to the 1992/1995 tab give rise to the same claim here – that is to say, the bottom radius which has to lie in line with the curve of the boss and the inner face of the tab. No other feature qualifies. In respect of this tab Qualtex claimed that the tab had to be wider to correspond to the extra width of the boss, which was itself widened to accommodate the screw. This was surmise on the part of Mr Hulme – it was not put to Mr Thomson. This chain of surmise led Qualtex to allege that the width of the tab had itself become caught by the must fit exception. This takes the section too far. It simply cannot be fitted in with its wording. Accordingly, the must fit exception is confined to the tubular fitting, the new boss design, the lower curve of the tab and the inner face of the tab so far as it retains the cable.

Must match. There is said to be a “closer match” to the tab on the Lower Hose Cuff because the tab is now wider. Whether or not this is a fair statement it still does not bring the tab within the must match provision for the same reasons as its predecessor failed. Qualtex also sought to rely on must match in relation to the extra width of the tab – it was said that the exception applied because it corresponded to the new boss design. Again, this does not get within the section – the relevant dependency does not exist. No part of the tab is excluded by must match.

Commonplace. In part the argument was a repetition of the argument that I have ruled fails in relation to the 1995 variant – the design was even more well-known in 1997 than in 1995. This argument fails for the same reason – being well-known is not the same as having become commonplace, and I hold that it had not. The screw attachment was also relied on as being commonplace. Inside the fixing is a spring which forces two indentations in the tube to co-operate with two ribs in the boss so as to encourage the winder to stay in the fully up or fully down position, while still allowing it to rotate – a sort of click mechanism. Mr Hulme said that screw fixtures for cable winders were commonplace and gave a historical example in the field of vacuum cleaners. I am not sure how compelling his example is, but he did not prove that this particular design existed in his example, so that evidence does not demonstrate that it is commonplace. Qualtex also produced what Mr Arnold described as “a rather ancient patent” (1936) relating to a vacuum cleaner which showed a cable winder biased against a spring. This shows at most that at that point someone had a similar idea. It does not show it was commonplace. The evidence on commonplace was insufficient to demonstrate that this fixing mechanism was commonplace. Last, Qualtex’s final submissions did not expressly rely on the general commonplace argument in relation to the shape of the tab, but for the sake of completeness I record that were it to be advanced, on the same material as Mr Hulme advanced in relation to the 1992/1995 tab, I would reject it for the reasons that I have given above in that context.

It follows that with the exception of the must fit aspects which I have referred to above, the 1997 cable winder attracts design right.

The DC01 1997 wand handle as a whole

The DC01 soleplate assembly

The nature of the item

In fact this item is made up of two separate items. They are shown separately in the second and third photographs of the part in the Annexe. The first of them is what is known as the cradle. It is held inside the other piece, which is known as the soleplate. The cradle is pivoted about two protruberances, one at either side, which fit into holes in the soleplate, and its movement about those pivots is partially restrained by a small rubber band fixed to the front of the cradle and which hooks over a hook on the front of the soleplate. This allows the cradle to adjust itself to irregularities in the surface being cleaned and to spring back when that adjustment is not required. The soleplate itself is fixed rigidly on to the bottom of the machine by screws. There is a piece protruding backwards towards the machine which contains an air channel; it passes up and fixes on to a correspondingly curved part on the bottom of the body of the machine. The large hook-shaped feature on the soleplate follows the profile of the rotating brush, and contains a channel down which the belt drive for the brush passes. I shall take the soleplate and the cradle separately and in that order.

The soleplate

DC01 soleplate - Originality

Mr Thomson told me, and I accept, that the starting point of his design was the former one piece design. However, the extent to which he copied that, as opposed to starting anew, was not fully investigated with him in cross-examination, and his witness statement does not deal with it. Looking at the piece, there are some aspects of it at the upper end of the air channel which resemble aspects of the fixed soleplate, but not many, (Footnote: ²) and many of those aspects are covered by must fit anyway, as will appear. Qualtex accepts that a number of aspects of the soleplate, and particularly those which relate to the cradle, were new in 1995. If there are any unoriginal parts they are not likely to be significant, or to affect the result of this case. In those circumstances I do not propose to address the subject of originality separately. I shall assume that the whole of this part was original.

DC01 soleplate - must fit

The rear of the cradle moves over a curved surface on the back of the soleplate. It is accepted that this interface is covered by the must fit exception. There is then a short vertical “wall” before the top of the back of the soleplate, which is also caught by the must fit exception. There is a dispute about this short wall – Qualtex says it is within the exception, and Dyson denies it. Mr Arnold’s final submissions did not expressly claim this feature, though his expert did, and I should deal with it. There was little debate about this in cross-examination. In my view it does not come within the words of the section. It does not enable any relevant juxtaposition.

The rearward extending part of the soleplate contains the channel up which the airflow passes. Qualtex claims that this channel is within the exception. Other than the actual aperture at the machine end, I find that it is not. It does not fit within the wording of the section – it does not “enable” a connection, or placing around or against either article within the meaning of the section. It enables the machine to work, but that is different.

Towards the top of the rearward extending portion there are two ribbed features with semi-circular cut-outs. These features are relied on by Qualtex as must fit since there is a spindle on the machine which sits in these cut-outs and around which the cleaner head revolves, according to the evidence of Mr Thomson. There is a matching feature on the cleaner head on the machine. This feature was not claimed in the pleadings, but it was claimed in Mr Hulme’s report. I am not sure whether in the end it was disputed by Dyson, but if it was then I formally rule that this feature clearly comes within the must fit exception. It enables the soleplate to co-operate with the spindle or spigot and with the cleaner head by fitting around the former and against the latter to enable all three to perform their respective functions.

It is accepted that the curved surface of the rearward extending portion is caught by the exception. However, there is a dispute as to one or both of the two side walls of that portion. In his final speech Mr Arnold claimed both those walls as must fit. His pleaded case claimed neither, but Mr Hulme claimed the outer wall (which is the far wall which cannot be seen in the photograph – it is the inner wall which is visible) which, he said, allowed the soleplate to fit and move against the wheel of the machine. I shall not deal with the inner, which was not part of the pleaded or expert’s case. As to the outer, there is a clearance of about 2-3mm between the wall and the wheel of the machine. It is not intended to interface in any particular way with the wheel, though (like any other part of the machine in that region) it obviously must not foul it. Looking at the two parts (wheel and surface), there is obviously no connection, and it cannot properly be said that the latter is designed as it is to enable it to be placed against or around the wheel. It is too far removed both spatially and in design terms for any question to arise as to that. It does not come within the must fit exception.

In his report Mr Hulme claimed that the inside of the soleplate had been designed to fit around the brush roll while leaving enough space for dust and debris to be sucked through the machine. He coloured a drawing which showed the parts which he intended to be within this description (along with other parts). In his final speech Mr Arnold claimed “generally” parts which fell within a similar description. I should make it clear that I will not make a finding in such general terms. Mr Jupp did not take issue with the pleaded case on this point (again achieved by reference to a drawing) and Mr Jupp agreed with that save for the vertical wall that I have referred to above. Because of the general way in which this point has been made I should make it clear that under this head I do not find that must fit applies to any feature of the soleplate which is not accepted by Mr Jupp in his report.

There is a rubber bumper strip which runs along the front of the machine at the bottom and turns down each side of the cleaner head. At the rear of the strip it is fixed into a gap between the “jaw” of the soleplate and the cleaner head. In his final speech Mr Arnold claimed features which enabled the bumper strip to be fitted to this gap (or recess, as he called it) and across parts of the side of the soleplate. This feature, or these features, were not dealt with in the pleadings or in either expert’s report. However, Mr Jupp admitted that where the bumper strip interfaced with the jaw of the soleplate and the opposite surface of the cleaner head there was an area which fell within this exception. Despite the fact that this was not pleaded or relied on by Mr Hulme, I propose to find that this area also falls within the must fit exception. Had Mr Arnold sought to amend immediately after this evidence was given I cannot see that the amendment could have been realistically opposed. In practical terms Mr Arnold just got on with the trial without making such an application, but that is not to be held against him. On this occasion, therefore, I will allow this point despite its not being pleaded or put in experts’ reports.

DC01 soleplate - surface decoration

DC01 soleplate - commonplace

Turning to the triangular ribbing, Mr Hulme produced three examples of ribbing in the context of vacuum cleaners. One was in a separate soleplate on a Hoover cleaner head, one was on the underside of a patented machine of which there was no evidence of commercial exploitation, and the third was in a Qualtex part (again a cleaner head) which was copied from an earlier part. The ribbing all differed in various ways, and to various degrees, from the ribbing on the bottom of the soleplate. None of them establishes that this ribbing was commonplace. They only show that the idea of ribbing (probably for strengthening purposes) had been used in three vacuum cleaners in various ways in the past. That does not even get close to establishing that the Dyson ribbing is commonplace.

I therefore find that nothing about the soleplate is commonplace.

DC01 soleplate - infringement

The DC01 cradle

DC01 cradle - originality

I find that the cradle was original in its entirety. There was evidence that the designers had seen the Wessel Werk piece, but it did not go much further. Mr Knox admitted that he saw the piece, and that he designed a piece which rocked in a similar way. Mr Thomson merely said he saw the part. The cradle on this piece does not look anything like the Dyson cradle. That is as far as the evidence in relation to other parts went. It therefore demonstrated a familiarity with an idea. That, or “partial inspiration” if that is the right description of that state of affairs, is not sufficient to deprive a design of its originality if it was otherwise created without copying, and there was no evidence of copying from the Wessel Werk product in the relevant sense. I find on the evidence that I heard that Dyson employees created this design without copying.

Two aspects (namely the front edge and bleed holes at the side of the cradle) were said by Qualtex not to be original over the earlier fixed soleplate. The evidence of Mr Knox on this point was that the front edge was “substantially similar” to the corresponding feature on the fixed soleplate, and the evidence of Mr Fitton (who assisted with the design) was that he accepted that he and his colleagues were “aware” of the bleed holes in the sides of the cradle. He did work to see what size and shape of holes worked best. It was not put to either of them that they copied prior designs and I find that they did not.

Accordingly the sort of copying which is necessary to prevent originality did not take place. I find that the cradle design was original.

DC01 cradle - must fit

In relation to the brush roll, it is common ground that the side features of the cradle which hold the brush roll in place are within the exception, as are the top sides of the two plastic strips which divide up the aperture through which the bristles rotate. Those strips stop the cleaner sucking up light carpets, and they have to be where they are because they coincide with gaps in the rows of bristles on the brush bar, and if they were anywhere else they would not allow the bristles to turn. In addition, Qualtex claims, as features which enable the brush roll to operate properly, “the whole of the inside of the cradle which is designed to fit around the brushroll providing clearance for the bristles and to define the suction path” (to quote from Mr Arnold’s written final submissions). This reflects one of the views of Mr Hulme, whose report said that the whole of the interior of the cradle was designed to fit around the spinning brush bar, its drive parts and end caps. It goes too far. There are certainly parts of the interior that are specifically designed to interface with the brush bar, and these parts are among those identified in the drawings annexed to the Defence and agreed by Mr Jupp and Dyson – for example, the rug strips, the curved section designed to accommodate the belt and the sides which interface with the end caps of the brush bar. However, it is not accurate to say that the whole of the interior is designed to fit round it. It is true that the interior should not foul the brush bar, but there is more space around it than is necessary merely to accommodate it. Part of that space is designed to achieve the necessary air flow. It is therefore not accurate to say that the whole interior is a feature which enables it to be connected to or placed around the brush bar in order to enable it or the brush bar to perform its function (to use the words of the section). In particular I reject the claim (if made) that the shape of the holes through which the bristles agitate the carpet fall within the exception. It is necessary for there to be some holes, but these particular holes were not designed exclusively to accommodate the bristles. They could have been larger, or they could have been smaller (but perhaps only slightly). As was pointed out in evidence, the size of the hole is affected by the desired airflow, and they might have been made a different size. The position of the fore and aft edges of the holes is not determined by the need to interface with the brush (save in the common sense way that they must not be so close as to foul it). I therefore reject this broad submission and that particular manifestation of it.

Next is a dispute about the lower front and trailing edges of the cradle. These are curved, and the evidence was that these curves assist the passage of the business end of the cleaner over the carpet. Qualtex says that that brings those aspects within the must fit exception; Dyson denies it. Dyson seems to take that view because (according to Mr Carr’s final submissions) only the flat portion at the bottom of the cradle engages the carpet (and as a result Dyson concedes must fit in relation to that flat underside). However, that seems to me to go contrary to Dyson’s own witnesses (Messrs Fitton, Knox and Thomson) who accepted that its purpose was to smooth the carpet pile out of the way. This means that these features enable the cradle to be placed against the carpet so that the cradle can perform its function (of passing smoothly over the carpet so that the brush and airflow can be directed so as to clean it). Since Dyson accepts that the flat parts are must fit, I find that these other features fall within the exception too. (I should record there was no debate as to whether the carpet was another article for these purposes – Dyson accepted that it was. )

The same applies to the shape of the chamfered edges of the front and back of the holes. The evidence of Mr Fitton and Mr Jupp was that these edges were shaped so as to interact properly with the carpet when the cleaner was being pushed and pulled. I find that these features too (in terms of their shape) fall within the exception.

Last under this topic are bleed holes in the side of the cradle. The sides of the large holes in the bottom of the cradle do not have flat bottoms. They have a rounded profile with two partially circular apertures which allow air to pass through them into the body of the cradle when the machine is switched on. Their purpose is to make sure that the airflow does not get completely blocked during cleaning, and to contribute to the airflow when cleaning up to something like a skirting board (according to Mr Fitton). Qualtex claimed must fit for these features. I do not consider that the exception applies to them. The essence of the must fit exception is that it applies to features that are essentially part of a connection – they facilitate a level of connection even if it is only to the extent of allowing one article to be placed against another. That is not true of these bleed holes. They are not part of the connection – if anything, they are an aspect of disconnection. They are a feature which enables the machine (and the cradle) to work, but that is not enough. They do not enable a connection of the kind referred to in the statute to be made. I can illustrate my point by taking the example of a modified electrical plug. Imagine a modern three square-pin plug which is to be used in a situation in which heat is likely to be generated at the connection with the socket. In order to dissipate this heat grooves are cut across the surface between the upper pin and the lower pins so that there is an air passage across this surface when the plug is in place – the plastic face of the plug is not contiguous with the plastic face of the socket along those grooves. Those grooves do nothing to enable the connection between the plug and socket – that works just as well without the grooves. They enable the plug and socket to be more effectively (and safely) used, but that does not bring them within the section. The bleed holes in the edges of the cradle are in the same category.

DC01 cradle - commonplace

DC01 cradle - infringement

DC01 lower hose cuff

DC01 lower hose cuff - originality

Must fit

The effect of this is that when the cuff is in place, at the back right the cuff lies along the plane of the trough. In the mid-section the ribs fit either side of the raised portion of the machine; and on the left the curve of the hose cuff does not follow the contour of the trough. Qualtex claims must fit for this entire back half of the cylindrical section; Dyson denies it.

I think that Qualtex is correct. It is clear that the right hand side is intended to fit into the corresponding arc on the trough. The ribs in the middle of the back of the cuff are intended to locate either side of the raised feature on the back of the machine even if the section between them does not follow the contour of that part of the machine. And the remaining third is intended to fit within the trough. Taking a realistic view of the section, I consider that the back part of the cuff is a feature which enables it to be connected to, or placed against, the back of the machine. It is a round peg which is designed to enable it to fit into a partially round and partially square hole, and if the fit is not perfect (which it is not) it is nevertheless a fit of sorts. I therefore find for Qualtex on this point – the rear curved face of the cylindrical section is excluded by must fit.

The position and shape of the part of the catch which fits into a correspondingly shaped hole in the back of the machine is claimed and accepted as must fit. Qualtex also claim the vertical tab on which that other part is mounted. I think it is right to do so. The raised feature at the end engages with the actual hole on the back of the machine, but it needs to be sprung into place and that is achieved by mounting it on a flexible tab. For these purposes the tab is a feature that enables the connection of the two parts and enables the raised part to fit be placed in the relevant hole and to stay there.

The next disputed area is the conical or flared section at the top of the cuff. Two aspects of this are accepted by Dyson as being must fit – the edges of that flare which abut the sides of the machine, and the rounded, beaded edge at the top. The former enables the cuff to be placed against the machine without rotation, and the latter enables the hose to be bent over the edge of the cuff (as it will be when used with a tool in the wand handle or at the end of the hose) without cutting through the hose. It struck me that in relation to this second feature it is not easy to fit it within the wording of the section. It is an interface (a phrase often used by Mr Jupp) but “interface” is not the right test. But be that as it may, that was Mr Jupp’s view and Dyson accepted that as its stance on this aspect and did not seek to resile from that. I will therefore proceed on the footing that that concession is correctly made. That leaves the inner (as opposed to the top) face of the flared part itself.

Mr Hulme considered that this flare was a must fit item. It was there to provide a “sympathetic curve” to prevent the hose adopting a sharp curve over the edge of the cuff which would imperil the integrity of the hose. On the facts, the hose can bend over the edge of the flare and, on many occasions, it will not touch the flare at any point other than at its beaded top (or where the cable winder is attached). It will only touch the actual surface of the flare if it is fully extended in its lower region; in that event some of the “rings” on the flexible hose will actually touch the flared surface.

The interior of the cylindrical part of the cuff is accepted by Dyson as being must fit. That is the part which attaches the hose, via a thread feature at the end. Although it is not threaded for its entire length, and although the hose is not physically in contact with the cuff for the entirety of the cuff’s length (there is a very small clearance from time to time) it is obviously right to treat the whole of this interior as being must fit because the whole of it enables the hose to be connected to the cuff so that each part can perform its function – the hose to function as a flexible hose, and the cuff to function as that part which joins the hose to the machine. The realistic and sensible view of the flare is that it does the same thing. It is in a very real sense an extension of the interior of the cylindrical portion. The interior guides the hose away from the fixing point in a secure fashion. The flare does the same at a spatially removed point. It enables the cuff to be placed against or around the hose so that the cuff can fulfil its function of protecting the hose, and so that the hose can fulfil its function of being a flexible hose safely. I know that I have just added the word “safely” to the statutory formulation, but the integrity of the hose is a quality of its function for these purposes, so it is merely elucidating its function, not adding a gloss to the statutory test. For those reasons I find that the flare falls within the must fit exception. The stepped portion immediately below the flare falls to be treated similarly for the same reasons.

Next there is the tongue/cable winder. This feature fulfils the same function as the winder on the wand handle – the cable is looped around it when the machine is not in use. The main dispute here concerns the shape of the vertical tongue. The cable fits against the horizontal “bridge” and is prevented from slipping off by the vertical tongue. Qualtex claims the whole of this feature as a must fit feature. Dyson does not in terms accept any of it, but Mr Jupp acknowledged that the cable would interface with the horizontal internal face and part of the internal face of the tongue. In my view the internal horizontal face falls within the exception in the same way as the horizontal component of the upper winder boss does. The outer face of the cuff, which retains the hose on one side, is in effect already within the must fit exception as part of the flare (see above). It is not possible to say that the whole of the inner face of the tab falls within it because the cable does not engage with the whole of the inner face; nor does it engage with the edges. The shape is irrelevant to the engagement. Quite how far the cable engages is not clear, but I do not need to indulge in a detailed consideration of that point because it is clear enough that must fit does not apply to the whole of the shape of the tab.

DC01 lower hose cuff - must match

As an afterthought to his first report, Mr Hulme pointed out that the cylindrical midsection of the cuff matched the cylindrical part of the upper hose cuff because they were the same diameter, and he thought that must match therefore applied. This is, with respect, fanciful. It confuses similarity or actual matching with the sort of design dependency that is required by the must match exception. There is no question of any such dependency in this article. If one changed the shape of the cylindrical section one would not necessarily radically affect the appearance of the whole assembly.

DC01 lower hose cuff –- commonplace

DC01 lower hose cuff - infringement

DC01 tool adaptor

DC01 tool adaptor - originality

Must fit

DC01 tool adaptor - must match

All those factors in the evidence are true, and if the question were whether the tube did in fact match the bulge then I think I could say that it did. However, that is not the test. The test is the statutory words, which do not contain the words “must” or “match”. The test is design dependency of the kind I have referred to above. I do not consider that this aspect of the design is dependent on the design of the bulge on the wand handle. The fact that it does match is not the test. If one applies the test of considering whether a consumer simply would not buy a differently designed part, I cannot believe that it would be fulfilled. Its design is nothing like as “dependent” as a car body panel. Mr Hulme said in cross-examination that one did not have to have ribbing but if you did then it had to match. I accept the first part without hesitation, and that is enough for this aspect of the case. It means the design dependency required by the section does not exist. I do not think I would accept the second part, but that does not matter.

DC01 tool adaptor - surface decoration

Were it not for the gripping function then I think that this would be a much more marginal decision on surface decoration than in relation to some of the ribbing. However, it is quite clear on the evidence, and I find, that giving good grip was a very significant, and certainly not a subsidiary, function. For that reason alone, therefore I find that this ribbing is not surface decoration.

DC01 tool adaptor - commonplace

DC01 tool adaptor - infringement

The DC07 stair tool

The DC07 stair tool - originality

The first DC07 piece was designed by Mr Howes. His instructions were to redesign the DCO3 head to improve pick up. He does not expressly say that he copied the ribs, but Dyson seems to accept that he did even though, to my eye, the profile of the ribs on the DC07 piece seems to be different and they do not appear to be an exact copy. I shall therefore proceed on the footing that they were copied. This means that the originality of this feature has to be tested as in the DC03 piece.

Mr Allard gave the history of the design of this piece. He admitted that he took another stair tool (the DC01 stair tool) as his starting point but it is obvious that he did not copy it. For certain of the details he took a brush tool as his starting point. In addition he agreed that he was influenced by the ribbing to be found on the brush tool. However, he denied copying those ribs and pointed out certain differences. The denial of copying was not challenged in cross-examination, and I find that he did not copy. Taking something as a starting point, or being influenced by something, does not make a product non-original. I therefore find that this ribbing was original in the DC03 part. It was copied over to the DC07 where it was not original, but its originality exists in the earlier part.

Other features of the DC03 tool were relied on by Qualtex as being non-original because they had origins in the earlier stair tool, and similarities were relied on. However, once again the evidence gainsays copying in the sense in which I have identified it above, so any relevant features of the DC03 tool which were carried over into the DC07 tool do not lack originality. I do not need to pursue the detail of this.

It was said by Qualtex that the shape of the head of the DC07 tool was a copy of the DCO3 tool and was therefore not original. That is accepted by Dyson – the shape of the overall head was copied in terms of its footprint. This is because the DC03 head was taken as the starting point to which modifications were made, and modifications were made to the upper surface and the internal parts of the underside. The footprint remained the same. However, this was original in the DC03 part.

The DC07 stair tool - must fit

In his final speech Mr Arnold claimed must fit for the tips of a comb that sits within the channel on the underside of the tool. It has six broad prongs whose job is to agitate the carpet fibres. In cross-examination Mr Hulme said that in hindsight he should have coloured this feature in as excluded on a must fit basis. That is too late to introduce this point, though I would add that the point is clearly wrong. The must fit exception does not apply to every aspect of a part that has a function (which it is likely to have) and which happens to lie against another part. There has to be an enablement of connection, or of placing, or of placing around. These combs do no such thing. They do not enable any of those things in relation to the part, or of any part of the part. They have a job once a connection is made, but their job is no part of the connection. Mr Hulme’s last minute attempt to refer to this is a good demonstration of his putting forward anything that might be argued to help his case rather than a properly considered dispassionate expert opinion.

Much the same sort of approach falls to be adopted to the claim made in the final speech that the internal channel in the head is a must fit item. Again, this was not claimed in the pleadings or Mr Hulme’s report, and it is not a claim that can therefore be entertained. However, I consider it is a wholly false claim. It simply does not come within the wording of the section when properly construed because it does not enable the connection or the placing on the carpet. It enables functionality, which is different.

Last under this head are various features said to enable the tool to fit on to machines for storage. Although this part is called the DC07 stair tool, it was intended that it be used with the DC04 machine as well. On the top of the cyclone part of that machine there is an imprint of the profile of the tool (when the tube is rotated so as to straighten it out) – it looks as if the tool has been pressed into the soft plastic and then removed. The purpose of the imprint is to hold the tool for storage when it is not in use. Qualtex claims “most of the underside of the head” as interfacing with this imprint. Dyson, through Mr Jupp, accepts that the shape of the outside of the head of the tool is within the must fit exception because it corresponds with that pressed shape. For my part I would have been sceptical about whether fitting into a storage compartment was within the section, but since it is accepted that it is I do not need to deal with it. The shape is covered by other concessions anyway. In his report Mr Hulme specifically claimed the backs of several of the ribs as falling within the exception for the same reason – that part of them fits within the groove. However, while the general outline may be within it (on the concession made) it does not seem to me that an area of the ribs is within it. If it matters I therefore find that the areas of ribs claimed by Mr Hulme are not excepted. The most that is is the outline formed by the outermost points of the few ribs in question. (And in fact the artificiality of this, to my mind, demonstrates the poverty of the argument that a shaped storage facility can bring “must fit” into existence in relation to a feature on the stored item which was never designed to facilitate storage.) Last, and briefly, Qualtex claimed that part of the flat underside of the tool was excluded because when stored upright on a post on the DC03 machine this flat surface almost abutted an adjacent handle. I regard this argument as fanciful and contrived, but it does not matter since the flat surface has already been conceded as must fit since it interfaces with the floor.

The DC07 stair tool - must match

The DC07 stair tool - surface decoration

On the front of the tool, immediately above the upturned leading edge, is a row of short ribs in the nature of buttresses which join the upturned edge to the body of the tool. These too were said to be surface decoration. (They did not feature on the DC03 tool). The evidence of Mr Howes was that they had a strengthening function, and this was accepted by Qualtex in Mr Arnold’s final submissions. This function was not subsidiary, so by itself it would rule out these ribs from being surface decoration. However, in any event, once again I consider them to be part of the overall shape and configuration of the article. It makes no difference that Mr Howes said that he added a radius to them in the DC07 version for visual purposes. That merely establishes that they go to appearance, not that that they are surface decoration.

The DC07 stair tool - method or principle of construction

Commonplace

The DC07 stair tool - infringement

DC04 hose assembly lower cuff

The DC04 lower hose cuff - originality

The DC04 lower hose cuff - must fit

Around the catch on the front of the cuff there is a cut-out area. Qualtex claims that the size and shape of this cut-out falls within the must fit exception because the size and shape of the cut-out determines the force necessary to depress the tab to clear the hole. Mr Jupp denied this. What matters is that the D-shaped catch is mounted on a tab. Having a cut-out isolates (and creates) the tab but the size and shape of the cut-out is irrelevant to this – it could be any size and shape. I agree with him. The tab falls within the exception because that enables the connection , but the size and shape of the cut-out area does not.

There is a dispute about the main part of the curved shape at the top of the piece. The dispute is much the same as the dispute in relation to the DC01 cuff. Mr Jupp and Dyson accept that the top of the curved part is must fit in relation to the hose, but they do not accept the cylindrical stepped out portion below it; nor do they accept the outer (or underneath) profile of the rounded portion. My conclusions on this are the same as in relation to the DC01 cuff – I find the inner cylindrical part and the surface of the curved “trumpet” (but not including the winder tab for these purposes – I deal with this below) are all must fit for the same reasons as in relation to the DC01 part. It does not matter that parts of the upper cylinder are not necessarily physically touched by the hose (which is true, but it is only true of parts). Other parts could be if the hose were very extended, and in reality the upper cylindrical part is part of an overall feature which comes within the section.

Last, there is a dispute as to the underside of the cable winder tab – the inner surface of the tab, the surface defined by the five ribs underneath it and the side body of the cuff opposite the tab.

As to the first, Mr Jupp pointed out that the vertical tab is not flat but is curved with an arc roughly parallel to the curve of the rest of the collar. Accordingly any cable wound round the tab would not actually touch most of the rear surface of this tab – it would only touch the edges. It was implicit in his evidence (though not explicit in his report) that he would accept the edges at least as being must fit. Qualtex would argue that that makes the whole of the rear surface must fit. I do not think that that is right. The feature of this design which enables the cable to be placed against (and secured by) this part of the cuff is the availability of the edges (or part of the edges – less than the entirety is necessary) of the cable winder tab. The rest of the curved surface is not there to do that. The tab is must fit only to this limited extent.

Mr Fitton said that the purpose of the ribs was to make a “nice surface” for the cable to wind round. It prevented the cable from snapping. Mr Jupp accepted that the cable wound around these parts. It seems to me that this inevitably brings the surface of those ribs within the exception in the same way as a continuous surface would have been. Qualtex do not claim more than the surface of these ribs under this head and I find they are right to do so.

The third side of the feature which retains the cable is the side of the collar opposite the tab. I find that this qualifies.

The DC04 lower hose cuff - must match

Mr Hulme also claims the shape of the winder tab as must match because it must match the upper winder. This, of course, mis-states the test. Furthermore, even if it were the test it cannot realistically be said to match. The upper winder has a curved end and straight sides which end at an annulus of greater diameter by which it fixed to the body of the machine. It is flat in cross-section, not curved, and it is differently sized and proportioned. If the bottom one had to match the top one, it would have failed. However, as I have said, that is not the test anyway. The design dependency required by the section is simply not fufilled. This claim fails.

The DC04 lower hose cuff - surface decoration

The DC04 lower hose cuff - commonplace

The DC04 lower hose cuff - infringement

The DC04 brush bar

A metal tube runs between these two end plates.

Over the metal tube is formed a plastic twist, or reverse helix, on which brush bristles are mounted. The helix reverses direction in the centre where there is a disc formed in the plastic. This disc is there to facilitate manufacture – it helps to support the brush bar in the bristling process.

Towards one end there is a pulley around which the drive belt fits.

The Modified DC03 bar had 53 bristle tufts along the bars of the helix. This was marketed in January 1998. It is this bar that contains the elements that have, in effect, been indirectly copied.

The DC04 original design. This was in substance the same as the DC03 (differences are dealt with below) with its 53 tufts. This was first marketed in February 1999.

The Modified DC04 brush bar design. This was marketed in mid-2001. It was the same as the original design save that it had 80 tufts and not 53. One of the tufts from the original DC01 design was omitted. This was the version actually copied by Qualtex.

The DC04 brush bar - Originality

The other attack on originality came from suggestions that certain ideas in the design came from ABC. In Qualtex’s final submissions it was said that a number of aspects of design came from Mr Hadfield and others at ABC. ABC applied bristles to the brush bars. This required ABC to drill out the holes on the flats of the helixes, and then fix bristles in them. Since they had considerable expertise in that area, Dyson employees had discussions with them about technical aspects the results of which were fed back into the designs. Sample brushes were created by ABC for testing by Dyson, and the final brush design was the fruit of this process. That much is common ground, and I so find.

So far as the giving of advice and the making of suggestions by ABC was concerned, reliance was put on the central disc, on two holes in the plastic (to provide a locating position for bristling), suggestions as to the thickness and constitution of the plastic and one or two other suggested improvements (mostly unparticularised). It was accepted by Mr Nighy that there were discussions, and I am prepared to find that the suggestion for the central disc did come from ABC (Mr Webber). Other things may have been suggested too. However, none of this makes the final design unoriginal. The suggestion of the disc was a suggestion of an idea, not the preparation of a preceding design. Accordingly, this evidence does not mean that Mr Nighy’s designs were not original for the purposes of design right under the Act, and I find they were.

There was a separate (though factually related) point in relation to the bristling patterns. In his witness statement Mr Hadfield said in terms that the bristle pattern for the DC03 (the 81 bristle pattern) was devised by him. He did not believe that Dyson initially supplied a bristle pattern for this, but merely sent the new moulding and asked ABC to bristle it. He would have drawn up the diagram to bristle the bar by hand or programme the bristling machine accordingly. Mr Nighy did not accept this. He said that the general pattern of activity as between the two companies was that Dyson would send sample bars down with bristling patterns attached, and that ABC would then bristle and return them. He denied that Mr Hadfield actually devised the bristle pattern. In cross-examination Mr Hadfield agreed that description of how things operated. Having heard that evidence, I think that Mr Hadfield is overstating the case. I think it likely that the 81 bristle pattern emerged after a period of trial and testing. Mr Hadfield might have given advice which led to the pattern being devised, tested and adopted (short of marketing) but there is no evidence from which I can conclude that he designed it first and Dyson employees copied it. I think it was ultimately a Dyson design. Mr Arnold urged on me the manner in which an earlier brush roll, the DC01 brush, was developed. This is a different style of brush bar, and is not in issue in these proceedings. I did not get any assistance from the circumstances of the design of that particular part and the function of ABC in that respect (which was again a matter of dispute).

I therefore reject this attack on originality too. The DC03 bristle pattern was original, as was the design of the rest of the brush bar. I should record that Mr Hadfield certainly did not claim to have designed the DC03 53 bristle pattern – he was most disparaging about it. However, nothing turns on that.

So far I have dealt with the bristling on the DC03 brush bar even though it is the DC04 bar that has been copied by Qualtex. The DC04 brush pattern is essentially the same as the DC03 pattern. One tuft has been taken out – a tuft on the central disc. There is one additional difference in that an outer tuft that was angled outwards has had this angle removed. Mr Nighy told me in cross-examination that the design for the DC04 bar was the “optimum” design for the DC03, a description which his witness statement gives to the 81 tuft pattern. The conclusion that I draw from this is that the DC04 pattern was not original over the DC03 pattern – it is in essence a copy. This means that as with the other features of the brush bar, design right dates from the DC03 bar (which is original in this respect) and not the DC04 bar.

The DC04 brush bar - must fit

In its final speech Qualtex sought to take a new must fit point. It sought to submit that all the parts of the brush bar were must fit because they were designed to be an integral part of the entire cleaner head, relying on evidence from Mr Nighy to the effect that the brush bar was designed to be an integral part of a complete drive system – without it the system will not work. I do not propose to allow this point to be taken because it was new in the final speech and probably a bit of opportunism. However, in any event it seems to be completely flawed as a legal proposition. To say it is part of a system is to say nothing relevant for the purposes of the must fit provision. The whole thing is not something that enables it to be placed in another article so that it (or the other) may perform its function in any sense apart from an unhelpfully generalised one. The end caps (to take an example) fall within the description, but not the whole thing. To say that it is designed to form part of a whole system is to say no more than that the design of the brush bar is intended to make sure that it performs its own job properly. Were it otherwise then many internal parts of many products which are intended to function internally would fall within the exception, which would be a very surprising proposition. The objections to this suggestion become clearer when one considers the case sought to be made in relation to some specific aspects.

Turning to those aspects, Qualtex rely first on the shape of the flange. This was not relied on in the pleadings or in Mr Hadfield’s expert evidence. I do not think that it is open to Qualtex to rely on it now. However, once again I will deal with it because it is apparent to me that it is a bad point. The flange is relied on in Qualtex’s final speech because “[it] was designed to support a reverse helix bristle pattern”. It was intended to allow the bristles to operate against the carpet without vibration. Mr Arnold also relied on evidence from Mr Jupp to the effect that something was needed to anchor the bristles; and it had to be a reverse helix bar to support a reverse helix bristle pattern. The fallacy in the argument is apparent from the formulation that I have quoted. If one imagines bristles in the precise helical pattern in this case somehow floating in mid-air, and requiring something to fix it, then the flange bar would be one way of capturing it and taming it. That, however, is nothing to do with the statutory “must fit” test. The test applies when one article is shaped so as to enable a fit with another. The bristle pattern is not such an article; it is a pattern. The bristles have a function – brushing the carpet – and the pattern may enhance it. But that does not make the pattern the article for the purposes of section 213(3)(b)(i). Nor does the existence of a flange, as opposed to some other configuration (such as a cylinder in which the pattern is fixed), “enable” the bristles (or indeed the pattern) to perform its function. This submission cannot be brought within the section.

Next in relation to the non-agreed items is the flanges on the belt pulley. In cross-examination Mr Jupp accepted in evidence that the inner faces of these co-operated with matching flanges on the soleplate cradle to keep dirt out. I find that that fact brings the inner faces of those flanges within the must fit exception – they enable the brush to fit in the cradle so that it may perform its function.

The next disputed item is the central disc. It is not necessary for the function of the brush as a brush. It is required purely for manufacturing purposes. According to Mr Nighy, it supports the brush when the bristles are being inserted, and it can be used to check alignment when the cleaner is assembled. Mr Hadfield explained that during bristling it sat in a V-shaped support on the bristling machine. Mr Carr accepted that the disc fell within the wording of the section if one treated the bristling machine as being the other article for the purposes of the application of the section – the disc was there to enable the brush bar to be placed on or in the machine so that the machine could perform its function. However, he disputed that the section applied to features which were there for manufacturing purposes rather than for the purposes of the machine when used. I think that he is entitled to rely on that distinction and that the disc falls outside the exception. The statutory provision is in my view looking at the situation of the article in its primary use, not at the stage of manufacture. It is a brush which is intended to be used in a vacuum cleaner, not an article which is intended to be bristled in a bristling machine. It is this stage that one looks at to see what its functions, or what the functions of a mating object, are. That is the context in which the statute is envisaging these exceptions operating, not the prior manufacturing stage. I therefore find that this disc is not must fit. I add that the use to check alignment was not relied on by Qualtex.

On each flange on each side of the central disc there is a point at which there is a gap in the bristles and at that point the helix briefly stops its twist and changes angle for a short stretch (about three bristle tufts in length) so that the flange is parallel to the central axis before resuming its twist. This gap coincides with the position of the rug strips on the bottom of the cradle. The bristle is omitted because if there were a bristle in this position it would act against the strip and not the carpet. This portion of the helix is relied on separately as a must fit item. Qualtex’s final speech erroneously records that there was a concession to this effect by Dyson. I cannot trace any such concession. Mr Jupp was asked to accept, and did accept, that the gap in the bristles was there to prevent interference with the rug strip. He was not asked about any “enabling” features of the change in direction. Mr Hadfield accepted that a mere gap in the bristles would achieve this purpose, and that the change in direction of the helix was to achieve a continuous pattern of bristle contact with the floor which would create an uneven sweeping pattern and the possibility of resonance. In doing so he was agreeing with the evidence of Mr Nighy to this effect. On this footing it is clear that this feature of the brush bar is not a must fit item. It does not enable the brush bar to be placed against the rug strips; it goes well beyond that and enables a full sweeping pattern notwithstanding a gap in the bristles at that point of the helix. It is therefore not covered by must fit.

The next point requires some further explanation of the construction of the brush bar. The central outer plastic core has a cylindrical metal tube running within it and bonded to it. Within that, and with a slight clearance, there is a metal tube. There is a bearing at each end of the brush bar which enables the metal/plastic core to rotate around this tube – the bearing is within the end of the metal core and surrounds the inner metal tube. Qualtex claims must fit in relation to (a) the inner metal tube (but not the tube), (b) the bearings and (c) the cylindrical outer plastic portions which cover the bearings at each end. Taking each in turn:

The bearings are clearly designed to fit around one article and inside another to allow rotational movement. That is what bearings of this kind do. If it matters at all, these bearings are within the exception.

I do not consider the outer shape of plastic (cylindrical in cross-section), just inboard of the end caps and in which the bearings sit, are covered by must fit. They do not enable the bearings (or anything else) to perform their function. They merely cover that which would otherwise be uncovered.

The bristle patterns are said to be excluded because they enable the brush bar to sweep the carpet and to perform correctly in the “integrated drive system” (to quote from Qualtex’s final submissions). The formulation of this submission was some considerable distance away from the statutory words, and further deliberation does not get it any closer. While it is true at one level to say that the bristle patterns enable the brush bar to sweep the carpet, one could also say the same substituting the words “whole cleaner” for brush bar (or even “human user”). It is not the test. The test is whether the bristles, or pattern, are a feature which enables the brush bar to be placed against (or in or around) some other article so that it, or the other article, can perform its function. The bristles (or pattern) do not fulfil this requirement. The function of the brush bar is to carry bristles and rotate them at high speed. The bristles (or their pattern) are not an enabling feature; they are the beneficiaries of the process. While it is true that pattern, length and other features affect efficiency, that is irrelevant for these purposes. Qualtex cannot bring itself within the wording of the section in this respect. The “integrated drive system” point has been dealt with above. I therefore reject this submission of Qualtex. It was also suggested by Mr Hadfield that the bristles are must fit because they are of a length that enables them just to reach the floor and “flick” the carpet. I am not sure what “feature” is being relied on here – the bristles themselves or their length - but whatever it is I do not think that it comes within the wording of the section. One cannot carry out any substitutional word changes within the section to produce anything which makes any real sense in this respect. I reject this suggestion too.

DC04 brush bar - commonplace

Qualtex also ran a commonplace argument in relation to the reverse helix bristle pattern. It was said by Mr Hadfield that the particular bristle pattern in this case (the reverse helix) was one of a number of commonplace bristle arrangements. Mr Jupp was unable to accept or deny this. It may well be that various shapes or ideas are commonplace, but it is not to those matters that the commonplace test has to be applied. It has to be applied to the Dyson manifestation of the idea – to the particular design in question. There was no evidence that this particular design, with its proportions, bristle tuft sizes and bristle spacing, was commonplace, and I find it was not. Dyson do not claim design right in the idea of a reverse helix bristle pattern (and indeed they could not do so in the light of section 213(3)(a)); they claim design right in their particular design which uses that idea. Previous deployments of helical patterns are not sufficient to make it commonplace.

DC04 brush bar - infringement

Large brush tool

Large brush tool - originality

There was an additional dispute as to originality in the bristle pattern. Dyson’s evidence, given by Mr Ford, was that Dyson sent over various patterns to be bristled as samples before deciding on the final bristle design. The samples were then considered, and the final design was decided on. In answers to me he said that the final design was dictated as much by aesthetics as anything else – whether it looked as though it would do the job. ABC and Mr Hadfield had some input in suggesting such things as bristle material. Mr Hadfield claimed to have devised the bristle pattern himself and produced a drawing which he had faxed to Mr Ford on 27^th October 1997 which sets out detailed positions and dimensions for the bristle positions. Dyson disputes that he devised it. Basing itself on Mr Hadfield’s evidence, Qualtex denies that the bristle pattern was original to Dyson. On this point I prefer Mr Ford’s general account. I find that various bristle patterns were proposed, and I do not doubt that Mr Hadfield had some significant input into the thinking of Dyson on those patterns because he was an expert and he would doubtless have suggestions to make. I think that it is likely that the pattern emerged as a result of trials and as a result of suggestions from Mr Hadfield, and its precise history cannot be ascertained from the evidence before me, but he did not design it himself in such a way that its adoption by Dyson would mean that they copied something designed by someone else so that their design was not original in this respect. The drawing to which I have just referred was, on the evidence, the drawing which Mr Hadfield created in order to set up the bristling machines, and he was sending it to his customer for the customer’s purposes. It was not some sort of original design which Dyson then copied.

I therefore find that the design for the large brush tool was original.

Large brush tool - must fit

This tool was supplied with, inter alia, the DC04 machine. When so supplied it is stored on the top of the cyclone part of the machine where there is a recess which is designed to accommodate the tube part. The machine is so designed that, when the head is rotated against the tube so that it is at an angle, the tube fits in the recess and what is in that orientation the bottom and the side of the head fit fairly snugly in a continuation of the groove and against the side of the machine. A rib on the machine is positioned so as to fit in the gap between the end of the head and the next adjacent rib. This method of storage is peculiar to the DC04. On other machines there is either no storage for it or it is stored on the machine in a way which does not involve the head engaging with any part of the machine. Based on this Mr Jupp accepts (and therefore Dyson accepts) that the shape of the tool head, insofar as it touches in this interface, is excluded by must fit. He also accepts that the groove between the rib and the endpiece of the head is excluded because it accepts the rib on the DC04 machine. I am not at all sure that I would have accepted that those aspects were must fit had it not been for this concession, but I have enough to decide in this case without revisiting areas which are conceded and I will not do so. However, Qualtex’s arguments take this further. Based on what Dyson concedes is must fit in this respect, Qualtex argued that the rest of the head fell within must fit “for reasons of symmetry”. I do not accept this – symmetry is no part of the must fit test – it simply does not come close to the wording of the statute. Qualtex also submitted that the rib and end part, which between them constituted the last groove on the head (or the first, depending on where one starts counting – I have started from the tube end) were excluded because they constituted the groove into which the storing rib fitted. I do not accept this either. Even if the groove is must fit, the particular shapes of the parts which constitute that groove are not within the wording of the section. They do not themselves, as a whole, enable the brush to be placed against the machine so that either the machine or the brush can perform their respective functions. The groove might do that and in the light of the concession I am prepared to find that it does, but nothing beyond that is within the concession or the statute.

Last under this exception, Qualtex argued that the rim of the head was excluded because it was a feature which enabled the bristles to be anchored in the head. It had to be of sufficient width and depth to anchor them. It is obviously true that, given certain bristles, if they are to be anchored in the head of a plastic brush, then the part in which they are fixed has to be wide and deep enough to accommodate them. That is almost a truism. However, it does not get Qualtex anywhere useful in this part of the debate. For something to fall within the must fit exception it must be a “feature … which enables the article to be connected [etc]”. To talk of aspects such as width and depth of an anchoring feature does not describe a “feature” for these purposes. It describes attributes of a feature. Attributes are not within the exception; features are. This particular head is a particular shape, which includes the attributes of its dimensions. There is nothing about this shape, as a feature, which falls within the statutory provision and it is not alleged that it does so – it was not alleged that this shape falls within it. Accordingly this aspect of Qualtex’s case fails too.

Large brush tool - must match

Large brush tool - surface decoration

Large brush tool - commonplace

Large brush tool - infringement

The DC02 wand handle

A plastic clip, which is an almost complete tapered hoop, which fixes the swivel on to the wand handle.

The body of the handle itself.

A bleed valve underneath the handle which is sprung against the user. If the user wishes to allow air to enter into the handle he or she presses the one end of the valve, most naturally with a finger in the same manner as a trigger is pressed or pulled. That depresses the finger end, which causes the other end to move away from the handle and admit air through the gap which is then created. Inside the valve there is a seal and a spring which do not feature in this case.

Must match – the whole DC02 wand handle assembly

I have reached the same conclusion in respect of this submission as I reached in relation to the similar submission made in respect of the DC01 wand handle. Of course the machine would look different with a different handle, but that does not go far enough to make it must match. This handle is not in the same category as car body panels (to take the familiar example again) for the same reasons as the DC01 handle, and this conclusion is even clearer in the case of this part because (as Qualtex conceded) this handle made a lesser contribution to the appearance of the machine as a whole. It is irrelevant that the particular substitute handle commented on by Mr Jupp is said to have a “dramatic” effect on the appearance of the machine, because it does not follow that any substitute handle would have that effect. If the degree of design dependency which I have held to be necessary is to exist so as to bring the overall design of the handle within the must match exception, then it would need to be the case that any alternative design would make the overall article (the machine) look radically different. The fact that one particular replacement has that effect (which is not what Mr Jupp actually said, in fact) does not mean that any other design would have the same effect. In any event, the terms of Mr Jupp’s acceptance is not the same as the test which seems to me to be applicable.

Furthermore, in the case of this part, there is some evidence from consumers which goes to cross-checking by reference to customer views. A substitute non-replica part is available for this handle, and Qualtex offered it for sale. More than that, it managed to sell it. (It is in fact the part that Mr Jupp referred to in the evidence just referred to). It is in essence an angled tapered tube with an appropriate diameter for the wand at one end and a fitting for the hose on the other. On top is a sliding bleed valve which would probably be operated by the user’s thumb. It is made of a grey plastic which is a completely different shade from the Dyson part. I was shown some purchase and sales figures for the period during which this part was sold by Qualtex as a replacement for the DC02 wand handle (1998 to 2003). Sales stopped when a replica part became available, and an order for the non-replica part was actually cancelled at that point (demonstrating that there was still some anticipated demand). The figures were part of a confidential exhibit, and it is unnecessary for me to set them out here (or in any confidential annex to this judgment) but it is sufficient to observe that in all but one of those years the sales were significant (in the hundreds) and their level demonstrates that there was customer demand for this definitely non-replica part. This is all the more so because these sales took place at a time when the machine was quite old and when one have expected the numbers of working machines (and therefore the demand for spares) to be falling and to be less than it would have been in previous years. Accordingly, if the test (or cross-check) for dependency is that a substitute product cannot be sold, then there is no dependency. If the level of customer demand is an indication of a commensurate lack of dependency, then there is no design dependency here. Accordingly this general must match point fails in relation to this part. A Mr Laybourne and a Mr Smith, both involved in the trade, gave some evidence about customer preference, but it did not demonstrate that a non-replica part would be unsaleable (though it supported a customer preference for replica items, which is insufficient for this purpose). Their evidence did not gainsay the clear impression given by Mr Hulme’s evidence of sales.

The various constituent parts of the DC02 wand handle

The wand handle

DC02 wand handle - must fit

DC02 wand handle - must match

DC02 wand handle - surface decoration

DC02 wand handle - commonplace

DC02 wand handle - bleed valve

DC02 wand handle bleed valve - must fit

I do not think that either aspect of this valve falls within must fit. The point on the rear portion fails on the facts. When the valve is depressed by the user, it does not actually depress sufficiently to enable it to lie flush with the sides of the handle, or so that the arc of the profile of the one lies exactly in line with the arc of the profile of the other. The valve stands slightly proud; and to my eye the two arcs have different radii anyway. That was Mr Jupp’s view, and I consider that he is right. It also fails conceptually because even if they were the same arc and even if they lay exactly in line when the valve was depressed, I do not think that this was a particular feature which enabled the connection, or placing against (or any relevant statutory conjunction with) the handle. It would be a design feature, but not enabling in a relevant way for the section. Again, this accords with Mr Jupp’s evidence, which I accept.

This latter point also provides the answer to the submission in relation to the front portion. In this area the arc of the valve is the same as the arc of the adjacent side of the handle, so that when the valve is at rest the two arcs are in alignment. That is true as a matter of fact. However, this feature is not enabling within the words of the section. It happens to be the case, and is no doubt designed deliberately that way. However, it does not enable the valve (or this part of it) to be placed in or against, or connected to, the handle so that either can perform its function. It is irrelevant to the interface between the two parts.

DC02 wand handle bleed valve - must match

DC02 wand handle bleed valve - commonplace

DC02 wand handle swivel

DC02 wand handle swivel - must fit

DC02 wand handle swivel - must match

DC02 wand handle swivel - commonplace

For once I think that this submission comes close. When the must fit items are excluded what is left is little more than a tube with an angled end. However, it still has the scallops and the internal webbing. Both Mr Jupp and Mr Hulme agreed that tubes with webbing in were common, because webbing is a common way of supporting an internal wall, but Mr Jupp said that this particular webbing was not. In the end I agree with Mr Jupp – what is left is not commonplace, but only just. The scallops are certainly not common, and therefore a tube bearing them cannot be commonplace.

The wand swivel retaining clip

The wand swivel retaining clip - must fit

The wand swivel retaining clip - must match

The wand swivel retaining clip - commonplace

DC02 wand handle - infringement

The DC02 - One outstanding issue

DC05 wand handle

Dyson’s design right case is confined to the collar and catch. The remainder of the wand handle is not original over the DC02 wand handle. No originality point arises in relation to those two features. I need not deal with Qualtex’s submissions which go to repudiate design right in any parts other than those two parts. I shall, in line with the general approach, take those two parts (catch and collar) separately.

DC05 wand handle catch - must fit

Must fit was claimed for the front surface by Mr Hulme in his report, but no claim was made in Mr Arnold’s final submissions for this surface. I cannot see how it can be – the top surface does not have to be its, or any particular, shape to enable it to be placed on or against any other article (or in any other juxtaposition relevant to the section).

DC05 wand handle - Must match

In order to deal with this point it is necessary to describe yet another part, namely the wand which fits on to the handle. The basic construction is two brushed metal tubes, one of which slides in and out of the other so as to give the wand varying length. At the point of the join there is a plastic locking feature, about which I need say no more than that. At the wand end there is a plastic feature which receives the end of the wand handle, and there is a shaped feature which receives the catch which I have been dealing with above. It has a raised perimeter which matches the shape of the end of the catch, and a notch underneath which receives the lip under the catch so that the catch holds the wand in place. At the other (tool) end there is another nozzle, (fashioned in plastic) on to which the tool fits, and there is a catch on that nozzle which is identical in shape to the catch on the wand handle and which engages with the tool in a similar way in which the catch on the handle engages with the top of the wand. Mr Jupp told me, and I accept, that the catches are probably fashioned from the same mould. The two catches are in line if one looks down the length of the wand.

Qualtex claims that the identical shape and apparently deliberate alignment gives rise to the must match exception for the shape of the handle catch. While it is apparently thoughtful design, I do not think that this brings the catch within the must match exception. I do not consider the design of the shape of the wand handle catch to be dependent on the design of the wand catch within the wording of the section. They are linked, but not dependent. Applying the approach that I have identified above to this section, the catch does not fall within it.

Nor does the allegedly chosen alignment of the catch with aspects of the collar on the wand end bring the shape of the top of the catch within the exception. On the facts the alignment is only approximate. The rear part of the catch (the part pressed with the thumb) sits in a similarly shaped aspect of the collar (whose shape at this point is conceded as must fit). It normally stands proud of that part of the collar. When pressed, the rear surface of the catch does not go down far enough for it to be aligned with the top of the collar in the sense that it is flush with it. It in fact stands proud (see above). The same applies to the other end of the catch when the catch is holding the wand. When the wand is in place at the end of the handle the end of the catch fits within a matching shaped feature on the wand (which makes the edge shape of the catch must fit, as to which there is no dispute), and it is lifted clear of it when the other end of the catch is depressed to release the wand. When the catch is within this feature it still stands proud of the matching feature on the wand end, not flush, and stands proud in varying degrees depending on which part of the catch one is looking at. There is no way that I can see that the design of the top of the catch is dependent on the design of this part of the wand as required by the section. This submission of Qualtex therefore fails. Once again this submission gives the game away by using the word “approximately” in relation to the allegation of alignment.

DC05 wand handle catch - commonplace

DC05 wand handle collar

DC05 wand handle collar - must fit

The inner circular face of the collar is accepted by Dyson as must fit (it interfaces with the nozzle end of the wand handle). The D-shaped hole underneath was not in terms accepted as must fit by Dyson, but Mr Jupp all but accepted it in his cross-examination, and so far as necessary I find that it is must fit because it enables the matching feature on the handle to perform its function of retaining the collar.

There is a dispute as to the angled or inclined surface which lies behind the catch. It is said to be must fit because it enables the catch to be comfortably depressed by the user’s thumb. Apart from being inconsistent with Mr Arnold’s disclaimer of a finger (or thumb) being half of a must fit argument, this point fails on the facts. When the catch is depressed the thumb does not naturally lie along the inclined surface – it may touch part of it (depending on how fleshy the thumb is) but no more. In my judgment the thumb-catch combination would still work in the same unimpeded way were the rear surface vertical or horizontal. The slope is not a feature which enables the thumb to perform its function. I therefore reject this submission.

DC05 wand handle collar - must match

The “approximate” alignment between the inclined rear surface and the depressed catch is relied on. This does not come close to creating the necessary dependency.

The ring part is said to align with the corresponding part on the wand or extension tube. If by the “corresponding part” Qualtex means the plastic end into which the wand handle is inserted then it is true that it aligns with the collar, but it is a mystery to me as to how this is said to make the collar a must match feature.

DC05 wand handle collar - commonplace

DC05 wand handle - infringement

PART 5 – COMMENCEMENT OF DESIGN RIGHT IN CERTAIN OF THE PARTS

Duration – commencement of design right

DC01

Prototypes for the DC01 were made in 1992. It was accepted by Dyson that orders for this machine were actually taken in 1992. By the end of December orders for at least 380 machines had been received. There was evidence that there were very serious discussions about further significant orders before the end of 1992. This level of interest is described in a contemporaneous memorandum as “trade commitment”, and it relates to at least 1200 machines, but it is not described as orders and there is no firm evidence that any more than 380 machines were actually ordered before the end of the year. In order to get these orders, and this level of interest, some machine must have been shown to potential purchasers, and I find that the prototype was. For these purposes the design of the prototype (then known as the DA0001) was the same as that of what was later called the DC01; Mr Carr accepted there was no material difference. However, no machines were delivered until, at the earliest, 12^th January 1993. BEAB approval was still being sought in December 1992, without which the machines could not be sold. It is unlikely that substantial production would have started without that approval (though it is clear that Dyson must have been all but ready to start in December). Looking at the available evidence on the matter I find that production machines were not produced until January 1993 – probably around the 6^th January, but the precise date in January does not matter. The first delivery took place in January 1993 and it comprised 440 machines delivered to GUS (probably comprising an initial delivery of 48 machines on about 12^th January with the balance following later, though again those details do not matter).

The position is therefore this. Actual production machines were not actually available (or even made) until after 1^st January 1993. However, something was shown to interested parties before that date, and orders were clearly solicited and, to a limited extent, actually obtained, before 31^st December 1992. These must have been obtained by showing the customers a prototype.

Qualtex say that in this state of affairs, “articles made to the design were made available for sale” before 1^st January 1992 because orders amount to making articles available for sale even if production had not started. Mr Arnold submitted that making an article available for sale can occur if articles to be made to the design are offered or exposed for sale without prior manufacture, but he also relied on the fact that some samples were available to show customers. Dyson says that the prototypes were not for sale, or available for sale, and that articles were not made available for sale until production started in January 1993. Pre-production orders do not fall within the section. At the point of production (but not before) real articles were genuinely available for sale; until then they were not.

This (like so many other questions concerning the statute which I have had to consider) is not an easy point, but in the end I consider Mr Carr’s case to be correct. There is apparently no known authority on the point, and no known authority on the meaning of “made available for sale”, which is the key wording in the paragraph. I agree with Mr Arnold that the paragraph is aimed at firing the starting gun at the point at which commercial exploitation starts, but I do not think that that takes the debate any further. The question is: what is that point? One could make a logical case for saying that in the context of the Act commercial exploitation starts when articles are offered for sale, whether manufactured or not; and one could make a case for saying that it starts when articles are offered having been made, or when they are first made after an offer. In other words, each side’s case is logically arguable. So I go back to the actual words used. I consider that the natural meaning of the expression “made available” connotes something that is actually in existence. If one imagines a case of an offer of goods which have yet to be made (in the sense that none of them are yet made) then I would not consider that those goods are “available” for sale even if advance orders for them are taken. Taking orders for them is not making them available. This is a common feature of today’s technology market. To take a modern example, the recently launched Apple mini-Ipod was heavily trailed in advance of sellers actually having any to sell (at least in this country), and orders were taken. I think that those who placed advance orders and who had to wait some weeks or even months to get their product might be a little surprised to be told that the product had been made available for sale when the product was announced and when first orders were taken. I do not consider that it is natural to say that a seller can make something available for sale when there is nothing there to sell. This conclusion is not affected by the existence of prototypes or samples. Any such prototypes or samples (on the evidence before me) were definitely not available for sale – they were examples of what would, in due course, be available for sale.

Applying that reasoning to the present case, I find that despite the advance orders taken in 1992, and despite the fact that a prototype was shown to intending purchasers (which for these purposes can be treated as an article made to the design), none was made available for sale until production started in January 1993. The prototype was just that – a prototype, and not something actually offered for sale, and there were no production models at the time which were offered (or available, to use the more relevant word). The facts of taking orders and having a made up machine available (I cannot avoid the word) for demonstration do not mean that articles are made available for sale within the meaning of the section.

Accordingly in the case of DC01 products in respect of which design right is claimed arising out of the first design of the product, the ten year period runs from the end of 1993; that is to say it expired at the end of 2003.

DC02

I do not consider that I can make those findings. There was no evidence from any mail order company about this, and such evidence as there was came from Dyson witnesses who were cross-examined about it (Mr Gammack and Mr Jupp). Their evidence on how these things worked was of some value as to general practice, but not nearly enough to enable to me draw conclusions as to how things actually worked in this case. The farthest Mr Jupp was able to go was that it “may be” the case that in the case of the DC02 there were advance sales but there were risks in making such sales which might have meant that there were none. When it was suggested to him that the photographs of the press launch meant that production must have started, he said that one could not necessarily draw that conclusion at all. There is no documentary evidence of any sales before March 1995, and certainly none that support a suggestion of sales in 1994. The suggestions to the contrary are mere speculation.

Accordingly, on the facts there is nothing to support any suggestion that the DC02 was made available for sale before the beginning of March 1995, and certainly not before the beginning of January 1995 (there is no practical difference between these two dates bearing in mind the wording of the statute and its reliance on calendar years). This conclusion would apply even if I had accepted Qualtex’s case that the date of orders was the material date. Accordingly I find that articles made to the DC02 design were first made available for sale in 1995.

DC03

DC04

PART 6 – PARTICULAR GENERAL DEFENCES

Acquiescence and estoppel

I shall therefore consider the facts in order to ascertain whether Dyson, knowing what Qualtex were doing, stood by and thereby encouraged Qualtex to carry on or thereby created expectations as to its (Dyson’s) future conduct and attitude in such a way and to such an extent as to make it inequitable for Dyson to insist on its legal rights now.

I have already pointed out that some spare parts for Dyson machines have been sold without any objection from Dyson – filters, drive belts and the brush roll for the DC01 machine are examples (though Mr Hulme thought that in recent times Dyson started to object to the last of these). Dyson’s part distributors are two companies called Electrue and Washvac. As well as dealing in Dyson manufactured parts, they also dealt in pattern parts made by third party manufacturers. They distributed both sorts of parts to other dealers. Mr Hulme thought that he and the pattern industry generally were entitled to produce replica pattern spares, but he knew that Dyson was litigious and thought that Dyson would or might sue even if they had no grounds. Accordingly, he says he decided to launch some parts, including the DC01 hose assembly in April 1998 and the DC01 tool adaptor in 1999, to see if Dyson would retaliate – “to test the water”, to use the phrase that he himself used. Dyson did nothing. He took Dyson’s silence as being an acceptance that they, like the other manufacturers, accepted that replica pattern parts could be lawfully produced. He therefore launched the other parts in the period from June 2001 (DC04 hose assembly) and November 2002 (DC01 wand handle). The last two parts that he introduced were the DC02 wand handle (September 2003) and DC05 wand handle (January 2004). There is some equivocation as to when instructions were given to the toolmaker for the DC05 wand handle. Qualtex’s pleaded case was that it was in about May 2003, which is after Dyson’s first complaint in a letter before action dated 7^th April 2003. In evidence Mr Hulme resiled from that and thought it was about March 2003. The cost of the tooling for these parts was not significant – less than “a few thousand pounds” according to Mr Hulme. He claims that it would now be unfair for Dyson to complain partly because he has suffered detriment in relation to his later tooling costs, but principally because had Dyson complained when the first products were launched he would have had the opportunity (which he said he would have taken) of withdrawing at that point and not exposing himself and the whole pattern spares market to the risk of adverse findings of principle (of the kind that I have determined above) that would make it more difficult for the market to carry on making duplicate spares. It is part of his case that Dyson knew about the launch of the earlier parts shortly after they were in fact launched.

In order for his case on acquiescence to get off the ground Mr Hulme must show that Dyson knew of the early parts that he introduced shortly after he launched them. Absent such knowledge he cannot make out any case of acquiescence or estoppel at all, since knowledge is an essential ingredient of this part of the case. Dyson’s case is that it did not know of the relevant competing parts until 2002. It knew about, and accepted, the pattern parts trade in such matters as filters, belts and the DC01 brush roll, but did not know about any of the parts which are the subject of this action until the time just mentioned. There is a dispute of fact in relation to this which I have to resolve.

Mr Hulme had no direct evidence of Dyson’s knowledge at this early stage, but he invited me to infer it from the evidence given by people in the trade that it was likely that the parts would have been drawn to Dyson’s attention via their main distributors within weeks of their launch. Mr Frank Smith has been involved in the spares business for 30 years. His evidence was that there was no secret about the Qualtex parts and they appeared in various flyers sent out to his customers. He had no doubt that they would have shown the flyers to Washvac and Electrue to try to get a better deal on Dyson original parts, and those two latter companies would be likely to have relayed the information to Dyson to get it to reduce its prices. Mr David Coombes, a consultant in the industry, expressed similar surprise that Dyson did not know earlier because the vacuum industry, particularly in relation to spare parts, is a close knit industry and in his view as soon as a new spare part appears on the market the rest of the industry knows about it. He too said that Electrue and Washvac would have urged Dyson to reduce its prices once the Qualtex parts came on the market. Mr Colin Beech now works for a domestic wholesaler but until 1995 or 1996 he worked for Electrue. He gave evidence of how it was that, in normal practice, new pattern products would be drawn to the attention of the relevant OEM in an attempt to get price reductions, and he explained how Electrue would have gone about that process. Thus Mr Hulme’s evidence is very indirect – it is based on how the market works rather than what happened in this particular case, and I am asked to infer that it worked in that way so far as his first two parts were concerned so as to give Dyson sufficient knowledge from which I can begin to infer acquiescence. He did not produce any direct evidence from Electrue or Washvac. He had attempted to get some third party disclosure from that quarter very shortly before the trial commenced, but he failed.

Dyson’s evidence was that it did not know about the relevant Qualtex parts until probably 2002. Mr Christopher O’Connor was the supply chain director from August 2001 until May 2002, when he became UK sales director. He left the business in 2004. (He was never a director in the company law sense.) In about May 2002 he carried out a review of the position in relation to sales and accessories and in the course of that he had a meeting with Electrue in which various pattern parts were shown, and after which Electrue produced a list of Qualtex parts. The market was then analysed and a report on spares was produced in July 2002. Attention was diverted by various other pressing business needs and pattern parts were raised at a meeting of 18^th September 2002, which involved lawyers. Mr Peter Richardson is now the Chief Operating Officer of Dyson, and had become managing director in January 2001. Prior to that he had been General Manager, Floorcare (June 2000 until January 2001), Group Sales director (August 1998 to June 2000) and UK sales director (October 1997 to August 1998). He had had overall responsibility for sales since October 1997. Until June 2000 he had direct responsibility for the UK sales team, and from June 2000 the UK sales directors reported to him. He explained that during the period up to at least 2000 Dyson’s focus was on producing machines rather than spares, and there was no real focus on spares until 2001 when a Mr Mark Mason was given responsibility for sales of spares and accessories. By this time this area had been identified as having potential for growth and in 2001 there was put in train a process to investigate the market in accessories and spare parts. He could not recall when he first knew that Qualtex was producing pattern spares, but he recalled being shown a document in the latter part of 2002 which demonstrated that fact. Mr Jupp remembers being shown a copy DC01 soleplate towards the end of 2001 but could not remember who it was made by or whether it was fixed or floating. He did not know about the Qualtex parts which are the subject of these proceedings until the latter part of 2002. There is no document disclosed in the action that puts the date of discovery of what was happening any earlier than 2002. The first communication explicitly on the subject was a letter before action dated 7^th April 2003.

Having considered the evidence of the relevant individuals carefully, I find that no-one in a relevant senior role in Dyson knew about the Qualtex parts that are relevant to this action before 2002. I accept the evidence of the Dyson witnesses that a review of the business, which led to a focussing on spare parts, took place at that time, and that in that context they were made aware of the emerging Qualtex parts (other than filters, belts and the DC01 brush bar which had been on the market for years and which Dyson will have known about and to which it did not object). The size of the business was not such that it would necessarily have been drawn to senior management’s attention until then, and Dyson’s business focuses were different, which may help to explain why they did not realise until then. If the Qualtex hypothesis was right, then Electrue and Washvac would have communicated the existence of new Qualtex parts to their account manager. From 1997 until June 2000 that account manager reported to Mr Richardson. In that period Qualtex launched 3 parts – the DC01 hose assembly, the DC01 tool adaptor, the DC01 soleplate assembly. Had the distributor raised a serious question about Qualtex parts it would have been raised with Mr Richardson. It is clear from Mr Richardson’s evidence that no such matter was raised with him, and equally clear that he did not pass any such matter on up the line (to the commercial director, to whom he reported). Accordingly, any discussions that there were between the distributors and Dyson were at account manager level, and they stayed there. I accept that there may have been some limited communication at the account manager level, but so far as it was, that is where it stayed. For reasons that will appear hereafter, that is not sufficient to give rise to the estoppel relied on.

Next is the requirement for some element of encouragement. There are no acts which have been proved which are capable of amounting to any level of encouragement. Dyson did nothing because senior management knew nothing until 2002. Mr Hulme has nothing except possible knowledge at a low level, followed by certain knowledge at a higher level in 2002, but knowledge by itself is not enough. For the reasons that appear below, he was not entitled to treat this as encouragement or anything like that.

If an estoppel is to arise, it must be demonstrated that the knowledge and the relevant acts of encouragement took place at an appropriate corporate level. It would not suffice to show that some lowly representative knew of the Qualtex parts. As a matter of principle Qualtex cannot rely on knowledge on the part of, and encouragement by, someone who would not be of an appropriate seniority to bind Dyson in this respect. The sort of things that Mr Hulme alleges he should be allowed to do would be the sort of things that would require decisions at a high level – certainly above account manager level. It is clear from my findings of fact that there was no actual knowledge or actual encouragement at senior level, and there is no reason for Mr Hulme to have supposed that there was an apparent knowledge or encouragement either. That is for this reason. Mr Hulme says that he wanted to test the water. He therefore put the parts out in the market, and in doing so incurred tooling costs of several tens of thousands of pounds. He was asked why he went about it this way, instead of writing to Dyson and telling them what he planned to do in advance and (in effect) invite them to try to stop him. The evidence clearly showed that he had participated in vigorous correspondence with Dyson in the past over different parts, and that he was prepared to argue his corner extremely firmly. His answer to that was that he “did not want to prod the giant awake”, and that he wanted to put some product out in the market, but not too loudly. He commented that if he had written to Mr Dyson or the CEO, that would draw attention by the written word to what he was doing, and that that was quite different from one of their distributors dropping an item in front of a Dyson representative and drawing attention to what Qualtex was doing. What he was in fact doing in this part of his evidence is demonstrating some equivocation as to the level at which he wanted to draw attention to what he was doing. He admitted that the natural reaction of Dyson would have been to have tried to stop the trade straightaway, had he written to Dyson, and that was why he did not do it. If he deliberately chooses a mechanism for doing all this which stops short of unequivocally drawing the facts to the attention of those with the appropriate seniority then he cannot be heard to say that he was justified in believing that senior management knew and approved of it all when in fact they did not. He took the risk of their not hearing about what was going on, and they did not. Accordingly, on the facts there was no knowledge or acts which might otherwise be encouragement at any sufficiently senior level, and Mr Hulme cannot properly be taken to have assumed that there was.

That is sufficient to deal with the estoppel argument, but there are additional points which are just as severe obstacles.

First, I am not satisfied on the evidence that, had Dyson complained about the first articles that Mr Hulme produced, he would have backed off and stopped producing them. His allegation that he has been deprived of this opportunity was an important part of his estoppel case, but in order to succeed he must satisfy me that he has acted to his detriment in not taking that course. That involves him satisfying me that he would have taken it. I am not so satisfied. He had spent £30,000 or £40,000 on tooling, and that would have been wasted. He held, and still holds, firm views as to Qualtex’s entitlement to produce the parts which he has produced hitherto, and in cross-examination he acknowledged the bad message that would have gone out to the rest of the industry if he had backed off. I think it unlikely he would have backed off. I think that he would have carried on and made Dyson sue him. I have reached this clear conclusion having heard how he gave his evidence on the point in the witness box, but it is borne out by what happened after he received the letter before action on 7th April 2003. He told me that he thought that that letter was a try on and Dyson were not serious and would not go through with any objection (which is at odds with what he said about his knowledge of Dyson’s litigious propensities). If that was his view in 2003, it would have been his view 3 or 4 years earlier. Production of the DC02 and DC05 wand handles was started in June and August 2003 respectively, which is after these proceedings were commenced. The latter was not sold until January 2004. So Mr Hulme carried on manufacturing in the face of the present action. That clearly indicates to me that Mr Hulme would have stood and fought whenever it was that Dyson chose to take him on.

For those reasons, too, the estoppel and acquiescence defences fail. I will not dwell on further weaknesses.

PART 7 - REMEDIES

Remedies

I was urged to find that Qualtex’s breach was flagrant. Dyson alleged that Qualtex set out to copy because it thought that it would make more money that way, and without any attempt to work out what parts were the subject of design right and which were not. It was capable of creating designs which did not infringe (the DC07 lower hose cuff and the DC04 upper hose cuff containing modifications) but generally chose not to. While there were minor differences on some of the parts, those differences did not mean that there was no copying.

For its part Qualtex relied on the bona fides of Mr Hulme and the fact that he honestly believed he was entitled to do what he did. Others in the industry shared that view, and the position on the law was not clear. This was the first spare parts case to reach trial, and a legal determination of the points arising was required.

I confess to finding it unsatisfactory to have to make findings of fact for the purposes of a future inquiry so that they can be used for the purposes of arguments which (presumably) were not fully deployed before me. There is a risk that I will not have dealt with all the points that might become relevant when full argument takes place, and even that the parties may seek to adduce further evidence of further fact on the points, or rely on points in the evidence before me in respect of which I have not made specific findings. It was not clear to me that either party was debarring itself from relying on further facts – I rather think that they were not. In the circumstances I shall largely confine my findings in this respect to matters of fact which one or other of the parties specifically urged on me, adding only matters that have already been the subject of findings in the context of the estoppel/acquiescence part of this judgment.

I therefore find as follows:

Mr Hulme’s thought processes were more general than analytical. He thought that the Act entitled him to do what he did without necessarily analysing precisely why in relation to each part of each article. It was apparent to me that many of the attempts to except many facets of the articles were arguments which were conceived for the first time in the context of this case, and they were not articulated (even to himself) by Mr Hulme before then.

Some of the arguments put forward (such as many of the “visual link” arguments that I have considered above) were fairly fanciful, and it is hard to consider them as reasonable. Others, however, were more reasonable – for example, the must match arguments in relation to the two wand handles was not a fanciful argument.

Mr Hulme embarked on his conduct with the attitude of a crusader, which attitude probably grew stronger over time. That does not, however, affect his bona fides.

The very high level of copying was deliberate in the sense that it was fully intentional, and was done because it was easier to copy, because it would produce a more marketable product and because Mr Hulme thought it was something he was entitled to do.

No arguments were addressed to me on the full meaning of “flagrant”, and I will make no findings as to whether what Qualtex did was flagrant for the purposes of section 229. If flagrant means deliberate copying irrespective of belief of entitlement, then what Qualtex did was flagrant. If it requires an additional element of actual knowledge of wrongfulness then it was not. If it contains an element of mens rea which would be satisfied by recklessness, then Mr Hulme came close if he did not cross the line in that I find that he was determined to do what he did even though he knew there was a real possibility that it would be held he was not entitled to do it. He was not prepared to try to clear it with Dyson in advance.

Mr Hulme sought to contrive a situation in which Dyson was manoeuvred into a situation in which it could not object to what he was doing in a somewhat sly fashion. I refer to my findings (and indeed his admissions) as to what he set out to do when he produced his first parts in 1998 and 1999 – he hoped to engineer a situation by producing products and seeing what happened in the hope that he could produce an estoppel rather than facing the issue in a more open fashion.

Dyson’s ultimate reaction was entirely foreseeable.

PART 8 - CONCLUSIONS

Conclusions

SCHEDULE

Parts agreed by the parties or accepted by their respective experts as being “must fit”

1992 cable winder

______________________________

______________________________

______________________________

DC01 Soleplate assembly

______________________________

DC01 Cradle

______________________________

DC01 Lower Hose cuff.

______________________________

DC01 Tool Adaptor

______________________________

DC04 Hose Assembly lower cuff

______________________________

DC03/DC04 Brush Bar

______________________________

Large Brush tool

______________________________

DC05 Collar and Catch

The underneath of the rear part of the catch which enables it to fit with the collar, including holes for the pivot pins.

The semi-circular rear shape of the catch.