This is the judgment of the court, to which all of its members have contributed.

This marathon litigation is about trade marks for electric shavers. The contestants are well-known major manufacturers of three-headed rotary electric shavers for the international market: Koninklijke Philips Electronics N.V. (Philips), a Dutch company for whom Mr Henry Carr QC appears, and Remington Consumer Products Limited (Remington), for whom Mr Simon Thorley QC and Mr Roger Wyand QC appear.

In October 2004 Remington won the latest leg of the UK marathon before Rimer J, who decided that the trade marks registered and relied on by Philips in this case were invalid.

In this appeal by Philips the main question is whether a Philips trade mark consisting of “the shape of goods” (the 452 Mark, as described and pictured below) is valid.

Remington’s challenge to validity invokes the functionality principle. The principle ensures that shape marks, like other trade marks, are indicators of the origin or source of the goods in question. They do not enlarge the scope of trade mark protection to cover technical solutions for the goods themselves. Other forms of intellectual property, such as patents, designs and copyright, are available for the protection of particular rights of defined scope and limited duration in specified aspects of the goods themselves. If trade marks were allowed to confer rights in technical solutions for the goods themselves, legitimate competition in the relevant market for the goods would be impaired. A shape mark registered in breach of the functionality principle would enable its proprietor to protect his products from the competition of other traders’ goods in a shape, which was necessary to obtain a similar technical result. Competition considerations prevail in the legislation, which implements the functionality principle. There are excluded from registration, prior to any consideration of the conventional trade mark questions of the inherent or acquired distinctiveness of a mark, signs consisting exclusively of the shape of goods, which is necessary to obtain a technical result.

There is no shortage of judicial opinion, either in the UK or elsewhere in the European Union, on the validity of shape marks for this type of electric shaver. The litigation between Philips and Remington is multi-national and at several levels of the judicial hierarchy. The most relevant judgments were those delivered in Philips Electronics NV v. Remington Consumer Products Limited by Jacob J in [1998] RPC 283, by the Court of Appeal in [1999] RPC 809 and by the European Court of Justice in [2003] RPC 2. Those judgments all related to the registration by Philips of another shape mark for electric shavers (the 208 Mark, as described and pictured below) and the alleged infringement of that mark by Remington’s DT 55 Wet and Dry Shaver. The Remington product was a three-headed rotary model. The three heads at the top end of the shaver were arranged within an equilateral triangle in a similar fashion to the 208 Mark. The outcome of those proceedings was that the 208 Mark was held to be invalid. It was absolutely barred from registration as a trade mark on the ground that it consisted exclusively of features of the shape of the goods, which were necessary to obtain a technical result.

The validity of the registrations of both the 452 Mark and the 208 Mark is governed by the provisions of the Trade Marks Act 1994 (the 1994 Act). The applicable provisions of the 1994 Act are derived from, and they must be interpreted, as far as possible, to be compatible with, the First Council Directive 89/104/EEC (the Directive) and with the rulings of the Court of Justice on the meaning of the Directive. The ability to register the shape of goods as a trade mark was introduced into English law by the combined effect of the Directive and the 1994 Act. Under the previous legislation the shape of goods could not be registered as a trade mark. Once shape marks became registrable in principle it was necessary to impose restrictions on their eligibility for registration in order to prevent the competition consequences mentioned in paragraph 5 above.

Litigation in other countries of the European Union between Philips and Remington about various trade marks equivalent to the 452 Mark has so far yielded a clutch of helpful judgments on this point from national courts: in Sweden (Swedish Court of Appeal 28 January 2004), France (Paris Court of Appeals 16 February 2005), Germany (German Federal Patent Court April 2004 and compare German Federal Court 21 April 2005 relating to a “device mark” of the kind described later in this judgment), Spain (Court of First Instance, Barcelona 6 May 2004) and Italy (The Court of Milan 26 February 2004). The foreign judgments speak with one voice, though in different languages: that Philips’ shape marks for their three-headed rotary electric shaver are invalid by reason of their functionality.

Philips appeal (with permission granted by Waller and Neuberger LJJ on a renewed application on 21 February 2005) from the decision of Rimer J on 21 October 2004.

Infringement proceedings were brought against Remington by Philips in respect of the sale of a variety of Remington models of three-headed electric rotary shavers with heads alleged to be identical or confusingly similar to the 452 Mark. Remington challenged the validity of the Philips trade marks by way of counterclaim. Rimer J made a declaration on the counterclaim that the registration of the 452 Mark by Philips is and always has been invalid.

The precise legal question for this court on the principal issue is whether Rimer J was wrong in deciding that the 452 Mark (TM No 1,533,452), which was registered by Philips under the Trade Marks Act 1938 with effect from 20 April 1993 in respect of electric shavers (class 8) and later re-entered in the register maintained under the 1994 Act (schedule 3), was invalid by reason of section 3(2)(b) of the 1994 Act. The subsection imposes a preliminary bar on the registration of functional shapes as trade marks for goods. Only if that preliminary obstacle is overcome is it necessary to consider the more familiar trade mark issues of the inherent or acquired distinctiveness of the mark.

Although Rimer J made findings of fact relevant to the distinctiveness of the 452 Mark, it was unnecessary for him to decide whether that mark was also invalid under sections 3(1)(b) or (c) of the 1994 Act. According to Philips a reference to the Court of Justice may be necessary before the objections to validity under those provisions could be determined against Philips by an English Court, certainly in the case of the device marks mentioned in paragraph 14 below. Mr Carr QC criticised the lack of clarity in the judgment of the Court of Justice on the questions referred by the Court of Appeal in the litigation over the 452 Mark. He submitted that the rulings required clarification in two respects: first, as to how the court should identify the “essential feature” of the mark and, secondly, as to the effect of the requirement laid down by the court that the essential feature is “only” or “solely” functional.

As he had held that the 452 Mark is invalid, it was also unnecessary for Rimer J to decide whether Remington had infringed it.

In addition to the 452 Mark the declaration of invalidity on the ground that the essential features of the mark are functional covers the Philips trade marks described in the proceedings as “the device marks.” Examples of them are pictured below. They were registered before the 1994 Act by Philips in class 8, either individually or as part of a series of marks under TM Nos 1 080 316 (the 316 Mark), 1 087 357, 1 124 415 and 1 203 652 (the Device Marks). They comprise two dimensional drawings depicting three circles arranged within an inverted equilateral triangle. The Device Marks, principally the 316 Mark, have been used in various ways: for example, on the packaging for Philips’ shavers and in advertising materials. Rimer J described the Device Marks as all consisting of “schematic representations of the head of a rotary shaver with three cutting heads” (paragraph 198). They are all registered as mark type “Device only.” That expression indicates that the mark is purely a drawing and that no word is attached to the mark. The first two are coupled with a disclaimer that the registration of the mark shall give no right to the exclusive use of a device of an electric shaving head. Although the use of a trade mark includes use otherwise than by means of a graphic representation (section 103(2) of the 1994 Act), Philips do not allege that the 316 Mark or any of the other Device Marks have been infringed by Remington’s shavers or packaging. Indeed, Mr Carr asserted that the Device Marks cannot be infringed by use in the form of the three dimensional head of the Remington shaver.

Rimer J’s decision that the registrations of the Device Marks were invalid was based on the grounds that they were registered in breach of sections 3(1)(b) and 3(1)(c), as well of section 3(2)(b) of the 1994 Act. They were functional, but, even if they were not, they were not distinctive.

A trade mark is defined in section 1(1) as

Section 3(2), which was enacted to implement Article 3(1) of the Directive, contains absolute grounds for refusing registration. Section 3 provides that-

The restrictions on registration incorporate the functionality principle and govern its application to the registration of the shape of goods as trade marks. As mentioned earlier, competition factors explain the restrictions. The purpose of the restrictions is to prevent shape marks from extending trade mark protection beyond an indication of the origin of the goods to conferring on the proprietor a monopoly in the goods themselves. This rationale for the absolute unregistrability of the functional shape of goods exercises a potent influence on the interpretation and application of the statutory provisions. It is such an important factor in the interpretation of section 3(2)(b) that we will expand our treatment of it to include discussion of the interpretation of the Directive by the Court of Justice.

The broad aim of Article 3(1)(e), second indent, of the Directive, which has been transposed into the similarly worded section 3(2)(b), was to preserve and promote competition in the market for the relevant products in respect of which the shape mark is registered. Trade marks, unlike most other manifestations of intellectual property, are renewable and potentially of unlimited duration as indicators of the origin of goods. They are not intended, however, to confer on the proprietor of the trade mark a monopoly right in other aspects of the goods themselves. The unregistrability of certain shapes is directed at the potential misuse of shape marks in order to combat competition from rival goods, which incorporate similar technical solutions and functional characteristics. If such restrictions on registration did not exist, a shape mark could be registered and used to create and conserve a monopoly in technical solutions for the goods in question.

As appears in [2003] RPC 2, the European Court of Justice ruled as follows on the interpretation of Article 3(1)(e) of the Directive:

We now turn from the general to the particular and consider the 452 Mark itself in its legislative and policy context. The overall shape of the 452 Mark is pictured below. It is a two-dimensional picture of the shape of the top portion of a Philips three-headed rotary electric shaver. The judge described the 452 Mark in these words-

The rival arguments focused on 3 elements of the overall shape: the clover leaf design, the raised rim and the lower residual areas of the faceplate. Reference was also made by Philips to small notional circular areas of the clover leaf surrounding each cutting head.

The essence of the case for validity, which was ably argued by Mr Carr QC for Philips, is that the 452 Mark does not consist exclusively of the shape of the goods, which is necessary to obtain a technical result. In short, Philips contend that the clover leaf is an essential feature of the 452 Mark, that it is an embellishment of the faceplate, that it is not functional and that the judge was wrong to hold otherwise. They also contend that the lower residual areas of the triangle were also an essential feature of the mark and were not functional.

We should record that it is common ground that, although the 452 Mark was registered before the 1994 Act came into force and is two dimensional, the validity of its registration is to be determined as if it were registered under the 1994 Act.

It is also important to bear in mind that, although there has been a lot of argument on the law, Rimer J’s decision on invalidity was founded on his findings of fact about the essential features of the shape and their functionality. He heard evidence from both sides given by experts, trade witnesses and members of the public. In his admirably clear and careful judgment he reviewed all the evidence in detail (see paragraphs 53 to 124).

In interpreting and applying section 3(2)(b) to the facts found by him Rimer J subjected to detailed analysis the judgment of the Court of Justice on the reference by the Court of Appeal ([1999] RPC 809) in the litigation between the same parties over the 208 Mark, which had been registered by Philips (TM No 1,254,208) in respect of three-headed rotary electric shavers. The 208 Mark pictured below was held to be invalid because it was functional and was registered in breach of section 3(2)(b) of the 1994 Act. Rimer J concluded that, for similar reasons, Philips were also absolutely precluded from registering the 452 Mark as a trade mark.

On the issue of validity the judge was obviously entitled to compare the 208 Mark, which had already been declared to be invalidly registered, and the 452 Mark, the validity of which was challenged by Remington in their counterclaim in these proceedings.

As pictured above, the 208 Mark consisted of the depiction of the head of a three-headed rotary shaver, which included shear plates for the three rotary heads arranged in the form of an equilateral triangle surrounded by a line.

On his comparison of the 208 Mark and of the 452 Mark Rimer J concluded that the only material difference between them was that in the 452 Mark

The crucial question was whether the addition of the clover leaf feature to the invalid 208 Mark made the otherwise similar 452 Mark a valid trade mark. Rimer J held that it did not. He reached that result on the basis of his findings of fact and on a structured approach to the issues raised by the interpretation and application of section 3(2)(b) in the light of the rulings of the Court of Justice on the Directive cited earlier and later in this judgment.

In his review of the evidence Rimer J first considered the clover leaf, which featured in the 452 Mark, though not in the 208 Mark, and the function performed by it. He summarised Philips’ contentions that only the three circular areas of the clover leaf immediately surrounding the cutting heads performed a technical function and that the rest of the clover leaf and the lower residual areas of the triangular faceplate performed no function, but were embellishments.

Rimer J correctly observed that in the litigation over the 208 Mark it had been held that the shape represented by it was necessary to obtain a technical result and that the only difference between that case and the present one concerns “ a shaver head in which the outer triangular line of the faceplate has been re-drawn in the shape of a clover leaf”(paragraph 125).

He went on to consider the evidence about the function performed by the clover leaf. He referred in particular to the evidence given by Mr Schiferli for Philips and by Mr Turner for Remington. He found as a fact that the whole area of the clover leaf performs a critical function in the technical shaving operation, namely stretching the skin and raising the hairs for effective and painless shaving (paragraphs 132 and 144). Philips have not appealed against that finding.

Philips have, however, appealed against the finding made by Rimer J in relation to the function of the lower residual areas of the triangular faceplate (paragraphs 133 and 144.) He accepted the evidence of Mr Turner for Remington that the lower residual areas in the 452 Mark contribute to achieving a smooth shaving operation. He said

His overall assessment of the functionality of the shape in the 452 Mark was that-

In its judgment the Court of Justice gave the following guidance on the scope of Article 3(1)(e) of the Directive:

In the light of that guidance Rimer J summarised his approach to the issues raised by section 3(2)(b) as follows-

The first step is to identify the “essential characteristics or features” of the shape of the goods in issue. Once they have been determined, the next step is to determine whether the essential characteristics or features consisting “solely” of the shape of goods are attributable only to the technical result.

As to the first step Rimer J decided that the clover leaf feature was not an essential feature of the shape represented in the 452 Mark (paragraph 135). He found that the clover leaf, as a feature of the overall shape of the goods, did not have “distinctive eye impact.”

Remington argued (as a ground in their respondents’ notice) that that finding was the end of the matter. That finding alone, that the clover leaf was not an essential feature or characteristic of the shape of the goods, should, they said, have resulted in a finding of invalidity at that point. There was no need for the judge to proceed any further or to consider the functional character of the clover leaf feature.

The position in this case was that the clover leaf feature was the only difference between the 452 Mark and the 208 Mark. The 208 Mark was held to be invalid in the earlier litigation. The 452 Mark was invalid for the very same reason that the 208 Mark had been held to be invalid. The addition of the clover leaf as a non-essential feature did not produce a valid registration. Remove the clover leaf and all that Philips were left with was the invalid functional 208 Mark.

Rimer J in fact went on to the second step and dealt with the functionality of the shape mark. He asked: was the shape of the faceplate as a whole necessary to obtain a technical result (see paragraph 140)? He found that it was.

Philips contended that the judge made three errors of law in reaching his conclusion that the 452 Mark was functional. They were that he failed to apply the correct test laid down by the Court of Justice for determining, first, whether a feature of the shape represented in the trade mark is essential; and, secondly, whether the feature of the shape was attributable “only” to obtaining a technical result.

The third error was in the judge’s findings of fact as to the technical effect of the lower residual areas.

We shall deal separately with each alleged error.

Did Rimer J apply the right test in deciding whether the clover leaf feature was an essential feature of the shape of the goods in question? The dispute is about how an essential feature of a mark should be identified.

Philips contended that he wrongly decided (paragraph 140) that the clover leaf was not an essential characteristic or feature of the shape represented by the 452 Mark. It was not, they argued, a separate feature of the 452 Mark. It was part of the overall shape of the goods. As is the case with determining questions of the distinctiveness of a mark, the judge should have looked at the overall shape and the impression it created and not singled out the clover leaf for separate consideration, asking whether it was noticeable to the public as a separate feature.

The criticism focused on the judge’s approach to the question of “essential feature” in asking whether the clover leaf feature was “a feature of the overall shape which has distinctive eye impact” (paragraph 135). He treated the clover leaf as a feature on its own. He then asked, as a matter of evidence, “what impact, if any, does the presence of the clover leaf have on the average user of shavers so adorned?” (paragraph 135). He decided whether the clover leaf feature was essential by reference to whether it was regarded as distinctive by the public on its own. He said that he was not satisfied that the overall effect of the evidence was that the clover leaf had a distinctive eye impact. He found that many people do not notice the clover leaf feature. They would not notice the difference between the 452 Mark and the 208 Mark. There was evidence that the focus of users was on the overall triangular faceplate, not on the shape which included the detail of the clover leaf (paragraph 136). After referring to the evidence he concluded that, although “it did not paint a wholly uniform picture,” the weight of it did not convince him that

Philips submitted that the judge, having wrongly looked at the component parts of the 452 Mark and selected the clover leaf as a separate feature, had then applied the wrong test in assessing whether a feature is “essential” in the sense indicated by the Court of Justice in the passage cited above. He had confused two different questions: the question of distinctiveness with the question of functionality. He had wrongly imported the notion of distinctiveness into an earlier stage and then had applied it inconsistently with the correct approach to distinctiveness under section 3(1)(b) and the proviso to section 3(1). The question of functionality was a preliminary question. It had to be determined and overcome before any consideration of inherent or acquired distinctiveness. Those questions would not arise, if the question of functionality were determined against the applicant. See Philips v. Remington [2001] ETMR 509 at paragraphs 12 and 16 of the Advocate General’s opinion.

Philips submitted, secondly, that any feature which contributes to the overall impression of the whole of the mark is an essential feature of the mark. The average consumer perceives the mark as a whole. He does not analyse its various details. The judge ought not to have excluded, as an essential feature, the whole of the clover leaf. He ought only to have excluded those parts of the shape that were trivial or de minimis in the sense that they would have no impact on the overall impression of the shape. The clover leaf itself could not be treated as trivial or de minimis. It had a considerable impact on the overall impression of the 452 Mark. The judge ought therefore to have concluded that the clover leaf is an essential feature of the shape shown in the 452 Mark.

We do not agree with Philips’ submissions on this point. In our judgment, Rimer J was entitled, on the evidence before him, to find as a fact that the clover leaf feature was not an essential feature of the shape in issue. He made no error of law in reaching that conclusion. He considered all the evidence, including the evidence from Philips that the effect of the clover leaf was to accentuate the shape, size and configuration of the cutting heads. He was nevertheless entitled to find on the evidence as a whole that many do not notice the clover leaf feature.

As in other areas of trade mark law the important factor is the impact of the mark on the eye of the average customer. The perception of the average customer for the goods in question does not depend on the dissection of the mark and on an examination of each feature of the mark. It turns on the feature which contributes most to the overall impression created by the whole mark. We do not accept Mr Carr’s contention that any feature which contributes to the overall impression is an essential feature of the mark. Not every feature of the mark has the same visual impact. The question of essential feature was one of fact and degree for the judge. He was entitled to find that the clover leaf did not have such eye impact as to make it an essential feature of the 452 Mark. His conclusion was consistent with, though not determined by, the judgments of the other national courts referred to in paragraph 8 above.

We also agree with Remington’s “short cut” submission that, once the judge had found that the clover leaf was not an essential feature, that for all practical purposes determined the validity of the 452 Mark and that he would have been entitled, in the circumstances of this case, to hold that the 452 Mark was invalid without proceeding to consider the arguments on the issue of whether the shape of the goods was necessary to obtain a technical result. It is very difficult in this case for Philips to escape from the consequences of the earlier decision that the 208 Mark was invalid on functional grounds. The 452 Mark was essentially the same as the 208 Mark, but for the clover leaf feature. As the clover leaf was not an essential feature of the mark, Philips had not acquired by registration anything more than they already had - an invalid registration of the functional 208 Mark. It follows that Philips’ appeal on the 452 mark must fail.

Although he decided that the clover leaf was not an essential feature of the shape represented by the 452 Mark, we think that Rimer J was sensible not to take the “short cut” and to deal with the technical result point in relation to the whole faceplate in the light of the guidance given by the Court of Justice in the case of the 208 Mark and its use of the words “only” and “solely” in the context of whether the shape of the goods was necessary to obtain a technical result. We too will deal with these points, although they are not necessary for our decision.

The judge said that there was no escaping the fact that the clover leaf formed part of the overall triangular faceplate and that the faceplate, as a whole, formed an essential characteristic or feature of the 452 Mark. The question arose as to whether the faceplate, of which the clover leaf formed part, was attributable “solely” or “only” to obtaining the technical result (paragraph 140). Those were the words used in the judgment of the Court of Justice. They do not appear in section 3(2)(b), which uses the word “exclusively.”

Rimer J concluded that the essential features of the 452 shape represented by the elements of the triangular faceplate were attributable only to the technical result obtained. He was not persuaded otherwise by the submissions of Philips that the clover leaf was not solely attributable to obtaining any technical result because areas of it, other than the three circular cutting head areas, performed no technical function; that the lower residual areas of the triangular faceplate served no function; and that those were aesthetic rather than functional (paragraph 141).

The judge sensibly observed that the Philips shaver was a piece of electrical equipment rather than a work of art and that its sole purpose was to perform a particular function: to shave in a speedy, convenient, comfortable and effective way. The clover leaf performed an essential function of stretching the skin and raising the hairs. As mentioned earlier, he found as a fact that the whole of the faceplate, including the clover leaf, contributed to the overall technical objective of giving the user a smooth, effective and comfortable shave. The whole of it was attributable to obtaining that technical result.

Rimer J rejected the argument that the ruling of the Court of Justice required that section 3(2)(b) be interpreted so as to require the court to analyse every part of each essential feature of a shape in order to see whether or not it performed a technical function. The correct approach was to consider whether the feature as a whole performed a technical function.

Thus he rejected the argument based on the use of the words “solely” and “only” in the judgment of the Court of Justice that, as the clover leaf feature was included in the design for aesthetic, as well as functional, reasons, it cannot be regarded as “solely” or “only” attributable to the technical result. That interpretation of the judgment of the Court of Justice was, he held, one which was likely to undermine the heart of the policy underlying Article 3(1)(e), second indent, as explained earlier.

Philips argued that, when Rimer J came to consider whether the shape of the goods was necessary to obtain a technical result, he did not give any weight or real effect to the words “only” and “solely” in paragraphs 83 and 84 of the judgment of the Court of Justice. The correct approach is that the prohibition on the registration of shape marks in section 3(2)(b) only applies if all the essential features of the mark are attributable only to a technical result. Rimer J should have held that the references to “solely” and “only” required every single part of every single essential feature to perform a function attributable to the technical result.

In summary, Philips’ case was that, as all essential features of the mark must be attributable only to a technical result for section 3(2)(b) to apply and as parts of an essential feature are non-functional, then it could not be said that the essential features of that shape are attributable only to the technical result. In such a case you are not looking “exclusively” (taking that word from section 3(2)(b)) at an essential feature of the shape, which is necessary to obtain a technical result. Further, it was argued that, if the mark contains a feature which is both functional and aesthetic, the paramount public interest in competition does not require that competitors should be free to use that feature of the shape to achieve the same technical result. The policy, Mr Carr submitted, was to enable competitors to use shapes of goods which are solely or only functional, not to use shapes which have eye appeal.

Like the judge we do not accept the submissions of Philips on this point. The judge was right to reject them. He correctly applied the law as laid down by the Court of Justice in the case of the 208 Mark in terms that are clear enough to make it unnecessary for this court to refer further questions to the Court of Justice on the interpretation of the Directive, even if the decision on this appeal depended on this point. Neither the terms of the judgment of the Court of Justice nor the policy underlying the provisions of Article 3 (1)(e)/section 3(2)(b) require the court to dissect the 452 Mark and to analyse each part in order to determine to what extent the role of that part is wholly or partially functional. Fragmentation of a mark is not the proper approach, either with functionality or with distinctiveness. In order to decide whether the shape of the goods in question is necessary to obtain a technical result the court must consider the mark as a whole. That is how the relevant section of the purchasing public would perceive it. A trade mark is not treated by the public or by the court as a mere assembly of component parts. The individual parts may, of course, be looked at for the purpose of making the overall assessment of the impression created by the whole. Taking account of the essential elements of the mark, an assessment has to be made as to its functionality. Having decided that the dominant impression of the 452 Mark was not created by the clover leaf, so that it was not an essential feature of the mark, the judge was entitled to conclude that the 452 Mark, like the 208 Mark, was in substance functional. The addition of the clover leaf to the 208 mark was an embellishment, which did not contribute to the overall effect or impression of the 452 Mark. The judge made an assessment of the functionality of the 452 Mark which was open to him on the evidence. It was not flawed by legal error.

Philips also challenged Rimer J’s findings of fact in relation to the technical effect of the lower residual areas of the faceplate. He had found as a fact that the whole surface of the clover leaf comes into contact with the skin and contributes to the overall technical objective of a smooth, effective and comfortable shave. Although Philips disagree with the judge’s conclusion, they accept that the finding was open to him on the evidence.

Philips have, however, challenged the similar finding made by the judge as to the technical result of the lower residual areas of the faceplate. It was submitted that his conclusion, which was based in part on the evidence of Remington’s expert witness, Mr Turner, was plainly wrong on the evidence before him.

The judge dealt with this point at paragraphs 133 and 144 of his judgment. He referred to the evidence of Mr Schiferli and of Mr Turner on the point. Mr Schiferli had accepted in cross-examination that the lower residual areas could come into contact with the skin during normal shaving, and that it was an important part of the design of the shaver that all parts which do touch the skin, including the lower residual areas, should be smooth. But he maintained his position that these areas provided no functional benefit. The judge went on to say this in paragraph 133:

In this passage the judge gave a number of reasons for accepting Mr Turner’s evidence. He said that it makes engineering sense. He referred to Mr Turner’s analogies. He referred to Mr Turner’s general proposition as to the application of force to the movement of a surface area (the faceplate of a shaver) across a deformable surface (skin). He referred to corroboration from the facts as regards model 68351. He accepted that model 9300 was successful despite having no lower residual areas but accounted for that as showing no more than another way of achieving a similar technical result.

Having given his reasons in that way for preferring the evidence of Mr Turner, and having then dealt with other points, the judge summarised the position at paragraph 144, concluding with the following passage:

Mr Carr argued that this finding of fact was not open to the judge on the evidence, as regards the lower residual areas. He pointed to the judge’s rejection of another aspect of Mr Turner’s evidence (at paragraph 131 of the judgment), and to Mr Turner’s lack of expertise specifically in relation to electric shaving, by contrast with the great experience of Mr Schiferli in that area, and he contended that the judge was wrong to find support for Mr Turner’s views in the facts relating to the 68351 model and not to find that they were undermined by the facts relating to the 9300 model.

It does not follow from the fact that the judge rejected Mr Turner’s evidence on one point on which he and Mr Schiferli disagreed, that he could not properly accept Mr Turner’s evidence on another point of difference between them. That point is made no more cogent by reference to the difference in the two witnesses’ respective experience. If Mr Carr is to succeed on this point it must be by reference to the details of the evidence, rather than on the judge’s preference of Mr Schiferli on another issue. He sought to do so in several respects, among them by seeking to identify the point on which Mr Turner’s evidence was rejected as being part of the basis for his opinion on the other point.

Mr Schiferli spoke of the function of the faceplate as being to facilitate stretching of the skin prior to the capture of the hairs, and acting as a reference plate to the surface of the skin. Mr Turner did not suggest that the lower residual areas had any stretching effect, but he said that they served to produce a smooth shaving action, as summarised in the passage quoted above from paragraph 133 of the judgment. His opinion was that a smooth shaving operation was better assured by a larger than by a smaller faceplate. As the judge said in paragraph 133, he supported that by analogies described in his expert report, with smoothing irons, and with trays as compared with tins, drawn under pressure across a carpet. The judge found these comparisons helpful.

Mr Carr sought to undermine that conclusion on the judge’s part by criticising Mr Turner’s evidence about the two other models mentioned. The 9300 model has no lower residual areas, but is said to produce a smooth shave. Mr Carr therefore said that this is inconsistent with Mr Turner’s theory that a smaller faceplate is less successful in assuring a smooth shaving operation. Conversely, the 68351 model was unsuccessful, he says, because the change of plane from the horizontal of the faceplate to the vertical of the handle was too sudden, and the edges were not sufficiently smooth at that point, not because there was no lower residual area. Thus, he submitted, this provides no support for Mr Turner’s theory about the desirability of a larger faceplate and the function of the lower residual areas.

Mr Carr also submitted that Mr Turner’s evidence as to the lower residual areas having a function in achieving a smooth shave was vitiated by his misapprehension that a raised rim round the cutting heads was a necessary or desirable feature of shavers with fully depressible heads. The judge accepted that Mr Turner was mistaken on the need for a raised rim: see paragraph 131 of the judgment where he said:

He did, however, go on, in paragraph 132, to observe that this difference between the two experts was more apparent than real, and to explain why. Nevertheless, Mr Carr argued that the rejection of Mr Turner’s evidence on the question of the rim undermines the judge’s reliance on his evidence as to the function of the lower residual areas in helping towards a smooth shave, because (as expressed in paragraph 44 of his report) that was itself predicated on there being an element of high friction produced by the raised rim itself. If there was no raised rim, there cannot have been the high friction which needed to be counteracted by the wider area of faceplate including the lower residual areas. That, Mr Carr contended, was supported by the comparison with the 9300 model and with the true reason for the failure of the 68351 model.

As regards the 68351 model, Mr Carr relied on the judge having said, in the course of his review of the evidence of Mr Asawa at paragraph 113, that the witness had agreed that, if the edges of the clover leaf are sufficiently smooth, there is no need to keep the lower residual areas. The point had been made to Mr Asawa by showing him the 9300 model, which did have smoother edges than the 68351 model, but had no lower residual areas.

Mr Thorley’s response to these contentions was to point out, first, that Mr Schiferli had accepted in cross-examination that the lower residual areas would or might come into contact with the skin in the course of ordinary shaving, at least on the part of some users, and to contend that, because of that, those areas do contribute to the overall function. They must, at least, be smooth, as Mr Schiferli accepted in cross-examination in this exchange:

Mr Thorley further contended that, although the judge rejected Mr Turner’s theory that a rim was necessary, in fact a rim is present in the relevant shape, because that is what the raised clover leaf feature provides, and that it is with the consequences of this feature that the judge was dealing in paragraph 133. If there were no raised clover leaf feature but a smooth triangular faceplate, then (quite apart from the fact that there would be even less, or nothing at all, to distinguish the mark from the 208 mark, which was held invalid in the previous litigation) the issue of high friction might well not arise. He also challenged Mr Carr’s contention that Mr Asawa in cross-examination accepted that the only problem with the 68351 was that it had sharp edges, and said that his evidence in chief that, if the transition from horizontal to vertical was to be rapid, it needed to be further away rather than closer to the clover leaf, was maintained in cross-examination.

Mr Carr in reply submitted that to say that because lower residual areas, if they exist, must be smooth, and to show that they are smooth, does not show that they are attributable to obtaining a technical result. They do not detract from that result, but that is not sufficient to show that they contribute to it. He also criticised Mr Thorley’s submissions in relation to the judge’s rejection of Mr Turner, as being based on an unjustified rationalisation of the grounds for the judge’s decision. As regards Mr Asawa’s evidence, besides making forensic play with a contrast between Mr Thorley’s submissions and a passage in paragraph 29 of the Respondents’ skeleton argument, he invited us to rest content with the judge’s own conclusion as to Mr Asawa’s evidence, as mentioned in paragraph 75 above.

Despite Mr Carr’s criticisms of the judge’s findings as to the functionality of the lower residual areas, we consider that the judge was justified in his conclusion on this point. The judge himself noted that in some respects Mr Turner’s evidence was not always easy to follow (paragraph 118) and clearly he thought carefully about the points of difference between the two expert witnesses. In so doing he had regard to the differences in their respective expertise, and to the fact that Mr Turner approached the issues from a non-specialised engineering standpoint, from first principles. Mr Turner’s analogies, with irons, trays and paint tins, might have seemed somewhat far removed from the processes of shaving, but it seems to us that the judge was entitled to derive benefit from them as he said in paragraph 133 that he did. As for the comparisons with the other two models, although Mr Carr can fairly say that the unsmooth edges of the 68351 model seem to have been its principal drawback, rather than the absence of lower residual areas, Mr Thorley’s citation of Mr Asawa’s evidence shows that it does also provide support for Mr Turner’s opinion as to the benefits of using a wider area, including lower residual areas. By contrast the success of the 9300 model, without lower residual areas, does no more than show that there may be successful designs which do not include lower residual areas. That does not of itself show that if a design does include lower residual areas, those areas do not perform any function. If they do perform a function, then they are attributable to achieving the technical result, and it would not assist Philips to show that there are other ways of achieving the result without the use of lower residual areas.

It seems to us that it is a legitimate comment on Mr Thorley’s part that, even though a rim is not functionally necessary, as Mr Turner had thought it was, nevertheless it is in fact present in the design under consideration, in that the clover leaf area is raised, and does provide a narrow rim round the outside of the shaving heads, to the extent of 270 degrees. On that basis the judge was justified in regarding Mr Turner’s views about the significance of a rim as relevant in this context.

Mr Carr is entitled to say that it is not sufficient to show that the lower residual areas do not detract from the functionality of the other parts of the design. Otherwise, as he said, a feature which undoubtedly did not have a function of its own, but was pure embellishment, could be regarded as functional unless it actually impeded the function of the other aspects of the design. But it seems to us that it was properly open to the judge to prefer Mr Turner’s opinion on the point, and to come to the conclusion that his view, that the lower residual areas did perform a function in assisting towards a smooth shave, was soundly based as a matter of engineering for the reason given in the third sentence in the part of paragraph 133 quoted above, namely “that the application of force to the movement of a small surface area across a deformable surface (such as skin) will be likely to meet greater resistance than the application of like force to a larger surface area”. It does not detract from that to observe that a particular model of a different design, without lower residual areas (such as the 9300 model), may also work well, and be commercially successful.

For those reasons we conclude that the judge’s finding of fact that the lower residual areas, as well as the clover leaf, were attributable to the technical result was open to him on the evidence. Accordingly if the determination of the appeal depended on this point, we would dismiss it for the reasons given above.

Rimer J held that the Device Marks were invalid, as registration of them was absolutely precluded by section 3(2)(b) of the 1994 Act. As with the 452 Mark, he held that their essential features were functional. Alternatively, he held that the marks were precluded from registration by section 3(1)(b) or (c). As explained above, the issue of functionality within section 3(2)(b) falls to be considered prior to, and separately from, the issues of distinctiveness.

Rimer J reached that conclusion on the section 3(2)(b) point “with a degree of hesitation.” He said-

Mr Wyand QC, who argued this part of the case for Remington, sought to uphold the judge’s declaration of invalidity as correct. He contended that each of the images in the Device Marks is a two dimensional representation of the three dimensional shape of the cutting head of a three-headed rotary shaver stripped to its bare essentials. Taking the 316 Mark as the principal Device Mark he said that it was a functional shape mark for the face of the Philips three-headed rotary shaver as much as the 208 Mark and the 452 Mark, with which the judge was entitled to compare the 316 Mark: it consisted of a triangular faceplate with an arrangement of three rotary blades inside it. Indeed, according to Mr Wyand, there was no feature of the 316 Mark that was not functional. It looked like the shape of goods in the form of the face of the head of the shaver. On this point he submitted that the judge correctly took into account the actual use of the marks in the sense of the public reaction to the shape of the 316 Mark.

Further, the lack of detail in the Device Marks did not prevent them from being a shape of the goods. Thus, said Mr Wyand, simply removing some of the detail from the representation of the shape of the goods in the 208 Mark, which had been held to be unregistrable as it consisted exclusively of the shape necessary to achieve a technical function, did not make the sign in the Device Marks registrable as a trade mark.

Although Philips had not sued Remington in this action for infringement of the 316 Mark or the other Device Marks those marks could, according to Mr Wyand, be used to prevent competition from all three-headed rotary shavers. Section 103(2) of the 1994 Act enabled the proprietor of a two dimensional mark to complain of infringement by use of the marks otherwise than by graphic representation of them. This would be contrary to the competition policy behind the exclusion of the functional shape of goods from registration as trade marks.

Mr Wyand submitted that the judge was entitled to take account of evidence from the general public to see what they thought the image showed. The evidence confirmed his view that the image was of the functional part of the electric shaver. The judge was correctly considering evidence of what the image showed. He was not considering distinctiveness, which fell to be considered at a later stage.

Philips argued that the Device Marks were not functional. They were figurative representations, each of them comprising a simple triangle containing three inner circles. Whereas the 452 Mark depicted the top portion of a three-headed rotary shaver, the Device Marks were two-dimensional abstract images, being highly stylised representations of shaver heads. They were clearly distinguishable from the appearance of the goods themselves. They were used on the packaging for Philips’ three-headed rotary electric shavers and they served as indicators of the origin of the goods.

A shaver made in accordance with any of the Device Marks would not, Mr Carr submitted, achieve a technical result at all. It simply would not work. They were not the shape of the goods and they did not serve any technical function at all. Section 3(2)(b) did not therefore apply.

Mr Carr referred to a certified translation of a ruling of the German Federal Court dated 21 April 2005 in the case of Rayovac Corporation v. Koninklijke Philips concerning the IR trade mark No 430 836 for razors. The picture is similar to the sign in the Device Marks. An appeal against the refusal of the German Patent and Trademark Office to withdraw protection was dismissed. The court stated

Mr Carr also submitted that it was wrong for the judge to consider evidence of the public recognition of the Device Marks by reference to their actual use. Evidence of actual use would be relevant to the question of acquired distinctiveness, but it was irrelevant to an objection to registration under section 3(2)(b). As already explained, functionality is a preliminary and absolute objection to registration which has to be determined prior to the question of distinctiveness.

On this point we fully understand the hesitation of the judge. It is a more difficult point than whether the 452 Mark is functional. With respect, however, we have arrived at a different conclusion that the Device Marks are not invalidated by section 3(2)(b).

The policy of the shape mark provisions is all important. Competition in the relevant market for the goods must not be impaired by the registration of the shape of goods, so as to prevent others from selling products incorporating technical solutions or functional characteristics.

There is no objection on that ground to the Device Marks as registered. They are all images with eye appeal. They are of an abstract, non-technical and non-functional nature. None of them is, in our judgment, a “shape of goods.” The images in the form of the three circles within a triangle, as shown in Device Marks, are shapes only in a general graphic and figurative sense. They are not the “shape of goods” in the functional sense used in the legislation. They fall outside the competition policy of the provisions and outside the scope of the functionality principle embodied in Article 3(1)(e), second indent, and section 3(2)(b) of the 1994 Act.

We would emphasise that no technical elements or details of the three-headed rotary electric shaver are depicted in the Device Marks. There is no question of the Device Marks conferring a monopoly of the functional features of the shape of the goods, as distinct from a mark which distinguishes the three headed rotary electric shaver manufactured by Philips from the shavers of another manufacturer.

As Mr Carr accepted, the Device Marks could not be used by Philips to prevent Remington or another manufacturer from making a three-headed rotary electric shaver. This is for the simple reason that the functional 3 dimensional shape of a three headed rotary shaver would not involve a “use” of the 2 dimensional Device Mark at all.

We also agree with Philips that the actual use of the Device Marks on the packaging and the matter of recognition by the public referred to by the judge in this context was not relevant to section 3(2)(b), which was concerned with whether the shape of the goods represented was necessary to obtain the technical result. Actual use was relevant to acquired distinctiveness, but not to the prior issue of functionality

Although Rimer J’s conclusion on section 3(2)(b) meant that the issue of distinctiveness of the Device Marks did not arise, he went on to deal with the point briefly. He held that he was not persuaded that any of the three marks enjoyed any inherent distinctiveness and he found that each mark consisted exclusively of a sign designating a characteristic of the goods concerned (paragraph 208). This was not a surprising conclusion given his view that the Device Marks were no less functional than the 452 Mark and the 208 Mark.

We are unable to agree with Rimer J on the issue of distinctiveness of the Device Marks. On this point the judge wrongly asked himself whether, when used on Philip’s packaging, the public would recognise the Device Mark as depicting a Philips’ shaver.

We can be brief on this point, as the authorities do not assist and the arguments are not elaborate. In our judgment it cannot be said that the Device Marks are exclusively descriptive or “devoid of any distinctive character” and so precluded by section 3(1)(b).

There was also ample evidence from witnesses on the issue of acquired distinctiveness of the 316 Mark (Mrs Billington, Mr Dunley and Mr Gilmore). It was irrelevant that the members of the public would recognise the Device Marks as depicting a Philips electric shaver, a functional product. Descriptiveness and distinctiveness are not mutually exclusive. A trade mark can be distinctive and simultaneously convey a description of the goods: see West (t/a Eastenders) v. Fuller Smith & Turner plc [2003] FSR 44 at paragraph 36.

We are also unable to agree with Rimer J on the application of section 3(1)(c). A mark could be distinctive and also convey a clear description of the goods. When used on the packaging for electric shavers the Device Marks could have a descriptive connotation, but the stylised presentation of the faceplate could also have a distinctive character.

The questions of the distinctiveness of the 452 Mark considered by Rimer J (paragraphs 149 to 170) do not arise nor do any of the issues relating to alleged infringement of it by Remington considered by Rimer J (paragraphs 171 to 197). It is unnecessary to reach any concluded view on them.

We do not consider that it is necessary in order to determine this appeal to make a reference to the Court of Justice under Article 234 for another ruling on the interpretation of the Directive. The ruling of the Court of Justice in the litigation on the 208 Mark has supplied this court with the necessary guidance in order to determine the validity of the registration of the 452 Mark and of the Device Marks.

For all the above reasons we dismiss the appeal by Philips relating to the 452 Mark. It is invalid by reason of section 3(2)(b).

We allow the appeal by Philips relating to the Device Marks. They are not invalid by reason of section 3(2)(b) or 3(1)(b) or (c).