England and Wales High Court (Chancery Division) Decisions

IN THE HIGH COURT OF JUSTICE
CHANCERY DIVISION
(APPEAL FROM THE UNITED KINGDOM
INTELLECTUAL PROPERTY OFFICE,
TRADE MARKS REGISTRY)

		Rolls Building 7 Rolls Buildings London EC4A 1NL
		01/10/2012

His Honour Judge Birss QC :

1. In 2004 Cadbury UK Limited ("Cadbury") applied for registration of a trade mark under application No. 2 376 879. The point of the application was to register the colour purple as a trade mark for chocolate. The Examiner objected to the application on the grounds that the trade mark was devoid of distinctive character. However, after Cadbury had filed evidence of distinctiveness acquired through use of the mark, the application was accepted and published in the Trade Marks Journal on 30^th May 2008.
2. The mark applied for was the following:

Mark



Description of mark:

The colour purple (Pantone 2685C), as shown on the form of application, applied to the whole visible surface, or being the predominant colour applied to the whole visible surface, of the packaging of the goods.

3. The rectangle above is a purple block when reproduced in colour.
4. The goods specified were to be

Class 30:

Chocolate in bar and tablet form, chocolate confectionery, chocolate assortments, cocoa-based beverages, preparations for cocoa-based beverages, chocolate-based beverages, preparations for chocolate-based beverages, chocolate cakes.

5. In August 2008 Société des Produits Nestlé ("Nestlé") opposed the Application in Opposition No. 97819. Mr Allan James, the Principal Hearing Officer on behalf of the Registrar of Trade Marks heard the opposition. On 20^th October 2011 he gave a preliminary decision essentially finding for Cadbury and allowing the application. In doing so Mr James amended the specification of goods in certain respects to reflect his findings on the evidence. He gave the parties time to make submissions on the amended specification of goods and on 2 December 2011 he issued his final decision allowing the application to proceed to registration.
6. The specification of goods which Mr James decided the mark would be registered for was:

Chocolate in bar and tablet form; chocolate for eating; drinking chocolate; preparations for making drinking chocolate.

7. Nestlé now appeals against the decision. Cadbury also filed a respondent's notice together with a cross-appeal on costs.
8. The matter came before me on 25^th May 2012. Simon Malynicz appeared for Nestlé instructed by RGC Jenkins. Emma Himsworth QC appeared for Cadbury instructed by Charles Russell. At the hearing it was agreed that the way forward was to deal with Nestlé's appeal first. Cadbury's cross-appeal could be addressed later if necessary.
9. Before Mr James, the issues included questions of principle and law but a substantial part of the dispute below related to the evidence concerning the use of the colour purple by Cadbury and its licensees, use of purple by others and the perception of third parties. Some of Cadbury's evidence of use had been filed at the examination stage and was adopted in the opposition proceedings and some had been filed during the opposition. Some also was filed in support of an earlier Cadbury application. Three Cadbury witnesses were cross-examined before Mr James.
10. On the issues of principle (and on a point on bad faith) Mr James found in favour of Cadbury. His conclusions on the facts, and having heard the evidence, addressed the various parts of the specification of goods sought separately as follows:

i) For "chocolate in bar and tablet form" he decided that the mark had acquired a distinctive character in the required sense (paragraph 112). There was evidence that purple had been used on Cadbury's Dairy Milk since 1914.

ii) For "chocolate assortments", although there was a general association between Cadbury and the colour that association was blurred by the use of other colours too. Cadbury's best selling assortment (Roses) is primarily light blue. The claim to long standing use of a consistently purple box for Milk Tray was not supported by the evidence and Cadbury's Heroes (which are in a purple box) was a relatively new product. Accordingly Mr James rejected Cadbury's case on "chocolate assortments", finding that Cadbury had not proved the colour was distinctive by the relevant date in that class (paragraph 114).

iii) As for "chocolate confectionery", he found that this term covers confectionery consisting of just chocolate itself and other confectionery products for which chocolate is an important ingredient. Focussing on the first group he noted that Cadbury's Buttons and Cadbury's Twirl had been sold in mainly purple packaging for more than 10 years. Mr James concluded on all the evidence that the colour was distinctive of Cadbury for chocolate per se (paragraph 115). However for chocolate confectionery consisting of a blend of chocolate with other things, he rejected Cadbury's case and so held that Cadbury failed on the category "chocolate confectionery" at large (paragraph 117).

iv) For the various beverages, Mr James noted there was no evidence that Cadbury's cocoa product had been sold in purple packaging (paragraph 22) but accepted that the mark was distinctive for drinking chocolate and preparations for drinking chocolate (paragraph 121).

v) For "chocolate cakes" Mr James did not accept Cadbury had shown the mark was distinctive at the relevant date (paragraph 123).

11. In order to reflect his finding that the mark was distinctive for chocolate per se but not chocolate confectionery at large Mr James allowed the registration for "chocolate for eating". The reason for the qualification "for eating" was to ensure the specification stayed within the original list of goods since "chocolate" alone would not do so because cooking chocolate had not been within the original specification.
12. Before me Nestlé does not seek to overturn Mr James' key conclusions arising from the evidence above nor do they challenge the finding on lack of bad faith. In particular Nestlé does not challenge the important findings that the mark applied for is in fact distinctive (subject to a point on categories of goods). The mark is capable of distinguishing the goods of one entity from those of another. However that does not mean that Cadbury necessarily must win. The main part of Nestlé's appeal is a point of principle. The argument is that the mark applied for is not a sign capable of being represented graphically and as such is not registerable as a trade mark. This point has a number of detailed facets which I will address below. Two secondary points are taken on the specification of goods (on the inclusion of plain chocolate and on the term "chocolate for eating"). Before me Cadbury supports the decision below. It submits the appeal should be dismissed. By its Respondent's Notice Cadbury argues that it would not obtain an unfair competitive advantage by registration of the mark. Neither side argued this as a free standing issue and I do not need to consider the Respondent's Notice separately from the general issues on the appeal.
13. Although the general point raises important questions of trade mark law which could be the subject of a reference or references to the CJEU, this court is not the final tier of possible appeal in these proceedings and both parties asked me to decide the issue and not to make a reference. Accordingly I will not do so.

The approach to appeals from the Registrar of Trade Marks

14. By section 76 of the Trade Marks Act 1994, an appeal from the decision by the Registrar may be brought to an Appointed Person or to the court. Ms Himsworth submitted that the nature and approach to such appeals had been helpfully summarised by Daniel Alexander QC, sitting as a Deputy High Court Judge in Digipos Store Solutions Group Ltd v. Digi International Inc [2008] EWHC 3371 (Ch); [2008] RPC 24 at paragraphs [5] and [6] in which he stated:

5. It is important at the outset to bear in mind the nature of appeals of this kind. It is clear from REEF Trade Mark [2003] RPC 5 (" REEF ") and BUD and BUDWEISER BUDBRÄU Trade Marks [2003] RPC 25 ("BUD") that neither surprise at a hearing officer's conclusion nor a belief that he has reached the wrong decision suffice to justify interference by this court. Before that is warranted, it is necessary for this court to be satisfied that there is a distinct and material error of principle in the decision in question or that the hearing officer was clearly wrong (REEF). As Robert Walker L.J. (as he then was) said:

"… an appellate court should in my view show a real reluctance, but not the very highest degree of reluctance to interfere in the absence of a distinct and material error of principle" (REEF, para. 28)

6. This was reinforced in BUD, where the Court of Appeal made it clear that it preferred the approach of the appellate judge but nonetheless held that there was no error of principle justifying departure from the hearing officer's decision. As Lord Hoffmann said in Biogen Inc v Medeva plc [1997] RPC 1 at 45, appellate review of nuanced assessments requires an appellate court to be very cautious in differing from a judge's evaluation.

15. Mr Malynicz did not disagree with Ms Himsworth's submission. I respectfully agree with the learned Deputy High Court Judge. I will approach this appeal on the basis summarised in Digipos. It may be noted, however, that save possibly on the secondary aspect of Nestlé's appeal concerning the correct drafting of the specification of goods, the issues arising on Nestlé's appeal are concerned entirely with points of principle.

The legislation

16. Section 1(1)(a) of the Trade Marks Act 1994 is as follows:

Trade Marks

1. - (1) In this Act a "trade mark" means any sign capable of being represented graphically which is capable of distinguishing goods or services of one undertaking from those of other undertakings.

A trade mark may, in particular, consist of words (including personal names), designs, letters, numerals or the shape of goods or their packaging.

17. Section 3(1) of the Act is as follows:

Absolute grounds for refusal of registration

3. - (1) The following shall not be registered -

(a) signs which do not satisfy the requirements of section 1(1),

(b) trade marks which are devoid of any distinctive character,

(c) trade marks which consist exclusively of signs or indications which may serve, in trade, to designate the kind, quality, quantity, intended purpose, value, geographical origin, the time of production of goods or of rendering of services, or other characteristics of goods or services,

(d) trade marks which consist exclusively of signs or indications which have become customary in the current language or in the bona fide and established practices of the trade:

Provided that, a trade mark shall not be refused registration by virtue of paragraph (b), (c) or (d) above if, before the date of application for registration, it has in fact acquired a distinctive character as a result of the use made of it."

18. These sections of the Act are intended to implement Art. 2 and 3 of the Trade Mark Directive (now Directive 2008/95/EC) which also correspond to Art. 4 and 7 of the Community Trade Mark Regulation (Council Regulation (EC) No. 207/2009).
19. In summary, so far as material for this case, to be registered a mark must be a "sign" and must be capable of being represented graphically (see s1(1) and s3(1)(a)). Nestlé's case is that the mark in this case satisfies neither criterion.
20. Before getting into the decided cases it may be noted that the potential problems arising from applications of this kind are not hard to see. Conventional trade marks such as trade names ("Cadbury") or logos (such as a glass and a half of milk on a bar of Cadbury's Dairy Milk) do not give rise to the same conceptual problems as what have been called "exotic" trade marks such as smells, colours per se and other things. The attraction of a trade mark registration is that provided it is used and the fees are paid, it gives a perpetual monopoly. The problem is the same as the attraction but from the other perspective. Unless the registration of trade marks is kept firmly in its proper sphere, it is capable of creating perpetual unjustified monopolies in areas it should not. This is not a new problem. In his judgment in British Sugar v James Robertson [1996] RPC 281, which related to whether the word "treat" was a valid trade mark Jacob J (as he then was) started his judgment with the well known passage from the judgment of Cozens-Hardy MR in 1909:

Wealthy traders are habitually eager to enclose part of the great common of the English language and to exclude the general public of the present day and of the future from access to the enclosure.

21. What was said a century ago about the great common of the English language might be said today about the world's great common of the visible electromagnetic spectrum. In that sense the case before me is really very simple. Can Cadbury, even if they have shown that the public associate the colour purple with Cadbury's chocolate, obtain a trade mark registration for that colour per se?

The cases

22. The cases which fall to be considered are a series of four judgments from the Court of Justice of the European Communities (CJEU) all concerned with more or less exotic trade mark applications. The first three overlap in time and it is helpful to see them in chronological order. The fourth comes later.
23. In 2001 the Hoge Raad der Nederland (the Dutch Supreme Court) referred four questions to the CJEU. The point was whether a telecommunications company (Libertel) could register the colour orange as a trade mark for telecommunications goods and services. The case was Libertel Groep BV v Benelux-Merkenbureau Case C-104/01. In his opinion dated 12 November 2002 Advocate General Léger said no. His view was that colours themselves should not be assigned for the exclusive use of certain economic operators and should remain available to everyone. He recommended that the CJEU should answers the questions posed by holding that Art. 2 of the Directive meant that a colour in itself with no shape or outline did not constitute a sign capable of being represented graphically and of being used to distinguish the products or services of one undertaking from those of other undertakings.
24. Meanwhile the CJEU was also considering another kind of unconventional trade mark, a smell. On 12 December 2002 the CJEU gave judgment in Sieckmann v Deutches Patent und Markenamt Case C-273/00. In that case the applicant had applied to register an "olfactory mark" which was essentially the smell of a compound called ethyl cinnamate. The CJEU decided that a smell could be a sign within the meaning of Art. 2 of the directive, holding that a sign did not have to be perceived visually, but to be a trade mark within Art. 2 the sign did have to be capable of being represented graphically. On the facts of the case the attempts to do so with a chemical formula, a deposit of a sample or a combination of those things did not satisfy the requirement of graphical representation.
25. The key part of the judgment in Sieckmann is in paragraphs 44 to 55. The upshot is as follows. The point of the requirement for graphical representation is to define the mark with clarity and precision. The precise subject of the protection afforded by the registered mark is determined by the graphical representation itself. The entry of the mark on a public register is there in order to allow access by the public and the competent authorities. A graphical representation must enable the sign to be represented visually by means of images, lines or characters so that it can be precisely identified. The graphic representation must be self-contained, easily accessible and intelligible. The sign must be perceived unambiguously and in the same way. It must be durable. It must be unequivocal and objective and avoid any element of subjectivity.
26. In May 2003 the CJEU gave judgment in the Libertel case. The court did not agree with Advocate General Léger's opinion, and held that a colour per se, not spatially defined, is capable of being registered as a trade mark. The judgment noted that to satisfy Art. 2 of the Directive the colour must satisfy three conditions: be a sign; be capable of graphical representation; and be capable of distinguishing the goods or services of one undertaking from another. A key part of the CJEU's judgment is paragraph 27 as follows:

27. In that regard it must be pointed out that a colour per se cannot be presumed to constitute a sign. Normally a colour is a simple property of things. Yet it may constitute a sign. That depends on the context in which the colour is used. None the less, a colour per se is capable, in relation to a product or service, of constituting a sign.

27. Following from this the CJEU considered the question of graphical representation based on its earlier judgment in Sieckmann (paragraphs 28 to 37). The Sieckmann criteria were summarised as a need for a description to be clear, precise, self contained, easily accessible, intelligible and objective. The court's conclusion was that a colour sample does not on its own constitute a graphical representation. A sample along with a description in words and a designation of the colour using an internationally recognised identification code may be considered to constitute a graphic representation. The Pantone reference in this case (Purple Pantone 2685C) is an example of an internationally recognised code.
28. Mr James summed up the effect of Libertel in the context of this case as being a judgment that colour may constitute a sign, but that it cannot be presumed to do so and whether colour constitutes a sign depends on the context in which it is used (Decision paragraph 71). I agree. Mr James went on in that paragraph to give examples of occasions in which colour would not be used as a sign (first using brown, as it is the natural colour of chocolate and second using a colour incidentally in an image on the packaging). They were presented as no more than examples and I agree with both.
29. Also pending before the CJEU at the same time as Sieckmann and Libertel was a reference from Germany about an application to register a combination of two colours (blue and yellow) "in every conceivable form". This was Heidelberger Bauchemie Case C-49/02. In his opinion delivered on 15 January 2004 Advocate General Léger invited the CJEU to review the position adopted in Libertel. His opinion was that colours and colour combinations "designated in an abstract manner and without contour in an application for registration" even if the shades were defined by a specimen and designated by a recognised classification, do not fulfil the requirements of Art. 2 of the Directive in the sense that they do not constitute a sign capable of graphical representation.
30. The CJEU gave its judgment on 24^th June 2004. Art. 2 of the Directive is addressed in paragraphs 15 to 40. It is clear that the court did not agree with Advocate General Léger that Libertel should be reviewed or changed. For example in relation to the crucial paragraph 27 of Libertel the court in Heidelberger said at paragraph 23:

23. As the Court has already held, colours are normally a simple property of things (Libertel, paragraph 27). Even in the particular field of trade, colours and combinations of colours are generally used for their attractive or decorative powers, and do not convey any meaning. However, it is possible that colours or combinations of colours may be capable, when used in relation to a product or a service, of being a sign.

31. The Court went on to emphasise (in paragraph 24) that for the purposes of Art. 2 the colours or combinations of colours sought to be registered must in fact represent a sign in the particular context of the case in order to prevent abuse of trade mark law and prevent unfair competitive advantages. Nevertheless it is clear that the fundamental legal conclusion of this part of the judgment in Heidelberger is that the decision in Libertel, that as a matter of law a colour itself is capable of being a sign, stands as the law. Whether a mark applied for is in fact a sign in a given context depends on the facts and circumstances but that is a different issue.
32. The court in Heidelberger went on to deal with graphical representation. On the general principles the court followed its earlier rulings in Sieckmann and Libertel (see paragraphs 25-32). Then the CJEU went on to consider the application of the law on graphical representation to the case before it. Paragraphs 33 to 35 state:

33. Accordingly, a graphic representation consisting of two or more colours, designated in the abstract and without contours, must be systematically arranged by associating the colours concerned in a predetermined and uniform way.

34. The mere juxtaposition of two or more colours, without shape or contours, or a reference to two or more colours 'in every conceivable form', as is the case with the trade mark which is the subject of the main proceedings, does not exhibit the qualities of precision and uniformity required by Article 2 of the Directive, as construed in paragraphs 25 to 32 of this judgment.

35. Such representations would allow numerous different combinations, which would not permit the consumer to perceive and recall a particular combination, thereby enabling him to repeat with certainty the experience of a purchase, any more than they would allow the competent authorities and economic operators to know the scope of the protection afforded to the proprietor of the trade mark.

33. The court dealt with some related issues (colour samples etc. in paragraph 36 and capacity to distinguish at paragraphs 37 to 39) and expressed its conclusion on Art. 2 at paragraph 40. The conclusion was:

40. Subject to the above, it must be accepted that for the purposes of Article 2 of the Directive colours and combinations of colours, designated in the abstract and without contours, may be capable of distinguishing the goods or services of one undertaking from those of other undertakings."

34. The fourth relevant judgment of the CJEU is in Dyson v Registrar of Trade Marks Case C-321/03. That case concerned applications to register the transparent collection chamber of a vacuum cleaner as a trade mark. Each of the two applications consisted of one or another version of a bagless vacuum cleaner made by Dyson with the written description: "the mark consists of a transparent bin or collection chamber forming part of the external surface of a vacuum cleaner as shown in the representation". The question referred related to Art. 3 of the Directive but in argument the issue of Art. 2 arose and the CJEU decided that it needed to be addressed as part of its decision.
35. In relation to Art. 2 the CJEU in Dyson reaffirmed its decisions in Libertel and Heidelberger that to satisfy the Article the application must satisfy three conditions – be a sign, be capable of graphical representation and be capable of distinguishing (judgment paragraph 28). At paragraph 34 the court reaffirmed its decision in Heidelberger that the purpose of the requirement to be a sign was to prevent unfair competitive advantage which would be an abuse of trade mark law.
36. In paragraphs 35 to 39 of Dyson the CJEU considered the particular case before it. The mark applied for was not a particular type of transparent collecting bin but was "in a general and abstract manner, all conceivable shapes of such a collecting bin" (paragraph 35). This did not satisfy the requirement of specificity (Sieckmann) since the mark could take on a multitude of different appearances (paragraph 37). This would give Dyson an unfair competitive advantage since it would prevent its competitors from selling vacuum cleaners having any kind of transparent collecting bin (paragraph 38). Accordingly the mark applied for was in fact just a property of the product concerned and was not a sign (paragraph 39 citing Libertel paragraph 27). The application was therefore contrary to Art. 2 and should be rejected.
37. These are the four judgments of the CJEU which have a key bearing on this case. In addition, I was referred in argument to further cases. Only two cases are worth mentioning in any detail at this stage.
38. In a decision dated 14^th November 2011 (Case 29 W (pat) 173/10) the German Bundespatentgericht ("BPG" or Federal Patent Court) considered an application to register a mark which was described in words as a "violet-purple filled, rectangular-like geometric figure" with a further detailed description of the figure, including a range of ratios of its length and width. Along with the description were images showing three versions of the bulging rectangular figure in question. The application had been refused and the court was hearing the applicant's appeal. The appeal was dismissed. The English translation of the judgment was not always entirely clear but the essence of the judgment is clear enough. To satisfy Art. 2 of the Directive the mark must be a sign, be capable of graphical representation and be capable of distinguishing (paragraph 1.2.1, based on the four CJEU decisions above). On graphical representation (paragraph 1.2.2) the court drew a distinction between an abstract colour mark and a mark like the one in question. For an abstract colour mark a linguistic description and an internationally recognised colour code designation would satisfy the need for graphical representation whereas the mark in question, which was a figure of variable length and width, did not satisfy that requirement. The mark applied for covered a variety of forms and was therefore not readily accessible and not understandable. It was therefore indeterminate and, following Dyson, did not satisfy Art. 2 (see also paragraph 2). The appellant sought to draw an analogy between the mark applied for and an "abstract colour mark" but this was rejected because in an abstract colour mark, although the colour does indeed appear in practice in different ways, the subject matter of the mark in question – the colour itself – is always the same (paragraph 1.2.2).
39. I take the upshot of the Federal Patent Court's decision to be that the mark before them covered "a variety of different appearances" and so, following Dyson, did not satisfy Art. 2, whereas an abstract colour mark, i.e. a mark consisting of the colour itself, would not fall foul of Dyson.
40. Finally I was also referred to BP Amoco v John Kelly [2002] FSR 5. In that judgment dated 2 February 2001, the Court of Appeal of Northern Ireland considered a trade mark essentially for the colour green on petrol stations. The judgment did not relate to the issues I have to decide and the case came before the CJEU cases I have referred to above. A point in the case arose concerning the true construction of the particular trade mark in question but that was a specific issue in that case.

The arguments before me

41. Nestlé's case is that the mark applied for in this case also falls foul of Dyson. Mr Malynicz submitted that the core principle of Dyson is that a trade mark which covers a multitude of visual forms, which are dependent on human subjectivity, impression and other factors, is not "a sign" but rather has the potential to lead to abuse and anti-competitive effects because of its inherent uncertainty. He submitted that this applies to marks of all kinds including colour marks. This argument addresses both the matter of whether the mark is "a sign" and whether it is capable of graphic representation.
42. In developing Nestlé's case Mr Malynicz made a number of further submissions which I will address in context:

i) That Cadbury's mark is not a sign;

ii) That Cadbury's mark does not satisfy the Sieckmann criteria for graphical representation;

iii) That more than one colour was implied in Cadbury's mark;

iv) That in this case a so-called "spatial delimitation" was necessary; and

v) That the concept of "context" in Libertel had been misapplied by Mr James.

43. A particular subsidiary argument was Mr Malynicz's submission that Mr James took a too narrow view of Dyson in paragraphs 73 and 74, submitting that here the Decision appears to confine Dyson to a point about marks which are functional concepts or some other property of the goods whereas in fact the principle is a wider one in that a mark covering a multitude of visual forms is not a sign under Art. 2. So the fact that purple plainly does not represent a simple property of chocolate (Mr James' decision paragraph 74) does not mean that there can be no problem under Art. 2. I do not think it is fair to read Mr James' decision in this way. It is true that in the context of Dyson he refers to the question of whether the mark applied for is a simple property of the goods but overall Mr James clearly saw that the principles arising from four CJEU cases included a question of whether the mark covered a multitude of visual forms (see e.g. paragraph 81 of the Decision). Whether Mr James derived that from Dyson does not matter. There is no error of principle revealed by this subsidiary argument.
44. Mr Malynicz's arguments (iv) and (v) above were summarised in his submission as follows:

"Libertel did not hold that a non-spatially delimited colour mark that specified a Pantone shade, would, in all cases, satisfy the 'sign' requirement. It depends on the 'context' and to date the only clear guidance on what this context includes is that of the Advocate General and the Court of Justice in Heidelberger"

45. I do not accept that submission. Plainly Libertel did not decide that a "non-spatially delimited" colour mark that specified a Pantone shade would in all cases satisfy Art. 2 of the Directive. But the court did decide the matter of principle that a colour itself was capable of satisfying Art. 2. Once the principle is established that a colour can satisfy Art. 2 the question becomes one of fact to decide whether in a given case it does so. But in this case Nestlé accept Mr James' finding on the evidence that the colour purple is distinctive of Cadbury for chocolate per se. (That acceptance is subject to a qualification which I will deal with later and does not matter at present.) It is true that Advocate General Léger's opinion in Heidelberger sought to import a requirement for a "spatial delimitation" or "delineation" of the use of a colour to be registered before a mark could satisfy Art. 2 of the Directive but this was put forward in his opinion as part of his reasoning why Libertel should be reviewed. The court did not follow that opinion. In my judgment Heidelberger does not assist Nestlé, it assists Cadbury. I reject points (iv) and (v). They are contrary to Libertel.

Nestlé's primary case

46. One answer to Nestlé's primary case might be thought to be that the existing case law makes clear that Libertel remains the law. Whatever the position may be relating to other kinds of marks such as transparent collecting bins (Dyson), variable figures (the decision of the German BPG) or indeterminate combinations of two colours (Heidelberger), a mark consisting of a colour per se (provided it is defined properly in words and with a Pantone code) is capable of satisfying Art. 2 even though it can obviously be used in an infinite variety of ways. Whether in fact it does satisfy Art. 2 will depend on the context (Libertel paragraph 27) but in this case, on the evidence purple is distinctive of Cadbury for chocolate.
47. In my judgment Libertel is almost but not quite a complete answer to Mr Malynicz's primary point and to his arguments (i), (ii) and (iii). If the mark in this case is indeed a mark of the kind considered in Libertel then it seems to me that the point has been firmly decided by the CJEU. Pure colour marks are in principle capable of being registered. To decide Libertel itself the CJEU had to go against the clear opinion of the Advocate General to the contrary and in Heidelberger the CJEU was invited to review that decision and plainly decided not to.
48. Ms Himsworth submitted that the distinction to be drawn was between an infinite number of uses of the same sign (such as would be possible with a pure colour Libertel mark) and an infinite or indeterminate number of signs (such as in Dyson, Heidelberg or the BPG decision). The former was acceptable while the latter was not. I must say I do not find that distinction as obvious or clear as Ms Himsworth does. To my mind the problem that the mark can be used in an infinite variety of ways seems really to be just the same for a pure colour mark as it does in the other cases mentioned, but it is clear that the CJEU has taken care not to take that approach. The section of the court's judgment in Heidelberger dealing Art. 2 takes care to distinguish between occasions on which it is referring to "colours and combinations of colours" and occasions on which it is referring to "two or more colours". Paragraph 23 of Heidelberger refers to "colours or combinations of colours" and clearly does not disagree with Libertel, which was concerned with a single colour. However the section on graphical representation in paragraphs 33 to 35, which refers to numerous combinations, is written specifically referring to a graphical representation of two or more colours. In Dyson the Court clearly had Libertel in mind and did not suggest that its reasoning there was intended to change the effect of its ruling in Libertel. A pure colour mark can always be said to be usable in a multitude of visual forms but I believe Libertel precludes that point from gaining traction since it is true of all pure colour marks.
49. However the fact that Libertel remains the law is not a complete answer to Nestlé's case on arguments (i), (ii) and (iii) because it is necessary to consider the particular mark being applied for by Cadbury in this case. As well as making clear that the mark is the colour purple with a given Pantone shade, the written description of the mark uses the expression "applied to the whole visible surface, or being the predominant colour applied to the whole visible surface" of the packaging of the goods. Nestlé submit that this language means that the mark is indeed indeterminate because:

i) First the phrasing presents alternatives (whole or predominant).

ii) Second the reference to "predominant" itself makes the mark subjective. To be predominant means the colour must be the strongest or main element (based on the OED definition) and to assess that would be impossibly subjective. Mr Malynicz submitted that the cross-examination showed that Cadbury's intention was for "predominant" to cover a very wide and subjective variety of circumstances. So even if Cadbury could be entitled to a pure colour registration for purple (Pantone 2685C) for the relevant goods, that is not the mark they have applied for.

iii) Third, the reference to purple being the predominant colour recognises that another colour or colours will be present and effectively means that the mark applied for is a colour combination mark and should be treated as such. If so it falls foul of Heidelberger since it is not spatially delimited, it is indeterminate and violates Art. 2.

50. I will address these three points in reverse order.
51. In support of the third point Mr Malynicz referred to the judgment of the Federal Court of Australia in Woolworths Ltd v BP plc (No 2) [2006] FCAFC 132 (4 September 2006) (Heerey, Allsop and Young JJ). In that case the court decided that an amendment to an endorsement to two Australian trade mark applications which related to the colour green for service stations was contrary to provisions in the Australian legislation prohibiting amendments which substantially altered the identity of the trade mark or which extended the trade mark rights (s65(2) and 65(5) of the Australian Trade Marks Act 1995). The amendment introduced the word "predominant" by providing that the colour green was "applied as the predominant colour to the fascias of buildings" (etc.). The court decided this change infringed s65(5) because after amendment the scope of the trade mark included the case where the green was to be applied as the predominant colour along with one or more other colours (judgment paragraphs 60 and 62). Mr Malynicz submitted that the same applied to Cadbury's mark and so it must be regarded as a colour combination mark.
52. Mr James rejected the third point at paragraph 78 of his decision (see also paragraph 72). He said:

78 […] the words "….applied to the whole visible surface, or being the predominant colour applied to the whole visible surface, of the packaging of the goods" serve to identify a specific use of the colour purple which constitutes a sign. That is not to say that the colour purple will not be used with other material on the packaging. It is clear from the evidence that it is often so used. However, that other material does not form a part of the sign applied for. It is not therefore appropriate to regard the mark as a combination or colours, or indeed a combination of the colour in question with any other distinguishing matter.

53. I agree with Mr James. A need for a systematic arrangement or spatial delimitation as discussed in Heidelberger does not arise as a result of the wording of the description of the mark in this case.
54. I believe the Australian Woolworth v BP decision was not cited to Mr James but I do not think it would make any difference. For one thing the context in which the Federal Court were considering the matter was entirely different from the position here. Whatever may be the position in Australia, in my judgment the wording of the description of this trade mark does not turn what is obviously a single colour trade mark into a mark consisting of more than one colour.
55. As part of developing the second point, Nestlé's case was that the amount of purple on the packaging would be variable, as would the position of the colour and the combination with other trade mark language such as the words Cadbury or Dairy Milk. Mr James dealt with this in paragraphs 80 and 81. He accepted that it is not clear from the mark how the colour would be applied to the goods or for example how much the advertising materials would need to be purple before they count as use of the mark. However as he said, the same criticisms could have been made if the application did not have the verbal description in question. I agree with Mr James. In my judgment this argument does not really relate to the particular written description of Cadbury's mark. It would arise just as much if those words were not present. Mr James held that if the point was a good one then it would not be possible to register a colour in the abstract and that would undermine Libertel. I agree with Mr James.
56. As regards the argument that "being the predominant colour" is too vague, Mr Malynicz relied in part on the cross-examination of Cadbury's witness Ms Watson from the Marketing Department. He submitted it exposed the vagueness of the trade mark. The point pressed was a question: did the minor role played by purple on packs of the Cadbury products Flake and Crunchie mean that those products were or were not covered by the trade mark in Cadbury's opinion? The answer was summed up by Mr James at paragraph 20 of his decision. Ms Watson did not know. Mr Malynicz said that the decision was too charitable as far as Ms Watson's evidence was concerned. Ms Himsworth invited me to read the whole transcript of the relevant cross-examination and re-examination. I did so. I can find no reason to disagree with Mr James' assessment of Ms Watson's evidence. In my judgment this point does not assist Nestlé.
57. Mr James addressed the vagueness of "predominant" in paragraphs 84 and 85. The word had probably been used with the aim of increasing legal certainty having regard to a particular view of the law at the time the mark was applied for, which is no longer the case (see paragraph 84), but in any event Mr James considered vagueness on its own merits at paragraph 85 and held:

85. Be that as it may, I do not consider that the words are too vague. It is true that they require a decision taker to make a judgment as to whether a colour is the predominant one used on the visible packaging for a product. Mr Creighton's evidence at paragraph 31 above gives a reasonable idea of how such a judgment would most likely be made. It does not appear to me to present more difficulty than other judgments required to be made in relation to infringement. For example, it does not appear to me to any more difficult than deciding whether consumers would regard the purple covered by the application as being the same as another purple, or whether such differences as there may be would be likely to go unnoticed by an average consumer.

58. Again, I agree with Mr James. In my judgment the word predominant used in this way in this mark does not introduce any more vagueness or uncertainty than is already present and acceptable in a trade mark registration of this kind. Ms Himsworth submitted that the wording in Cadbury's mark simply made explicit what was implicit in the mark considered in Libertel. I agree. The colour will be the predominant colour applied to the packaging of the relevant goods.
59. Mr James' paragraph 85 refers to Mr Creighton's evidence. In order to mitigate what might be thought of as the uncertainty in Mr Creighton's evidence as well as the point about Ms Watson's cross-examination (above), Ms Himsworth submitted that the scope of an expression like predominant was not a matter for evidence at all. I agree but in any case I do not think that is what Mr James was saying in paragraph 85. He was not relying on Mr Creighton's opinion at all.
60. Mr Malynicz also submitted that Mr James had failed to take into account the fact that there had always been consistent use of purple by third parties. I reject that submission. Mr James considered the evidence of use of purple by third parties at some length (decision paragraphs 24 to 36 including a four page table summarising the evidence) and he took into account third party use when considering the distinctive character of the colour on the various goods (see paragraphs 101, 107, 115 and 119).
61. As regards the first point, that the description has alternatives (whole or predominant), Ms Himsworth submitted that these were not alternatives in any impermissible sense because one was simply the extreme case of the other. Covering the whole surface was simply the extreme case of being the predominant colour. Mr Malynicz's answer was say that even if the one is just the extreme of the other, they are still alternatives and referred to an analogy about claiming the letter A or every letter of the alphabet. I did not find the analogy compelling. On this point I think Ms Himsworth is right. The description in this case is not presenting alternatives, it is claiming the use of the colour as the predominant colour. That use may consist of coverage of the whole visible surface or it may cover less, provided it is the predominant colour.
62. Nestlé submitted that the wording in the description of Cadbury's trade mark cannot be ignored. I agree. Nestlé submitted that when the wording is considered, this case can be distinguished from Libertel. That is the issue I have considered at length above. For the reasons given, I disagree. In my judgment the wording which forms part of Cadbury's trade mark does not have the effect that this case can be said to be any different from the single colour mark considered in Libertel.

Conclusion on the main part of Nestlé's appeal

63. Since single colours per se are, as a matter of European law, capable of being signs within Art. 2 (i.e. they are capable of being a sign, capable of being represented graphically and capable of distinguishing) then, to paraphrase a little the words Cadbury have used in the description of the mark, in my judgment the colour purple (Pantone 2685C) applied to the whole visible surface, or being the predominant colour applied to the whole visible surface, of the packaging of chocolate, is capable of being a sign within Art. 2.
64. The answer to the question I posed at the outset is clear on the basis of the current state of European law. Since on the evidence the public associate the colour purple itself with Cadbury's chocolate, Cadbury are entitled to a registered trade mark for that colour on the relevant goods and that is the mark they have applied for. Mr James was right to find that the mark satisfied Art. 2. I will dismiss this part of the appeal.

The specification of goods

65. Nestlé submitted that on his findings Mr James should only have allowed the following specification of goods (shown with Nestlé's proposed insertions in italics and proposed deletions struck through):

Milk chocolate in bar and tablet form; milk chocolate buttons chocolate for eating; drinking chocolate; preparations for making drinking chocolate.

66. The two issues are whether the chocolate should be limited to milk chocolate and therefore not include plain, dark or white chocolate and whether "chocolate for eating" was acceptable as a category. If not then the only alternative category would be milk chocolate buttons.

Milk or plain chocolate

67. Nestlé submitted that the goods should have been limited to milk chocolate. First there was no evidence that Cadbury had used the mark in relation to plain, dark or white chocolate and second it is unsustainable that the evidence supports a finding of acquired distinctive character in relation to plain chocolate.
68. Mr James addressed this issue in paragraph 128 of the decision as follows (omitting a footnote reference):

128. I recognise that plain chocolate, milk chocolate and white chocolate are recognised types of chocolate. And it is true that Cadbury only uses the colour on the milk chocolate products that make up the vast bulk of its business. Further, Cadbury's own witness, Ms Watson, makes the point that the distinctiveness of the colour has been built around products for which milk chocolate is the key ingredient. On the other hand, all types of chocolate share the same basic properties. And I have found that a significant proportion of consumers would expect chocolate sold under the mark to be a Cadbury product. I therefore reject Nestlé's submission on the […] point.

69. However Nestlé submit that the finding that a significant proportion of consumers would expect plain chocolate to be sold in a purple package had no basis. Cadbury's case is the decision, based as it was on the category "chocolate" as a whole, was fair and was open to Mr James on the evidence.
70. I think a finding on the correct and fair level of generality in a specification of goods, by the Principal Hearing Officer of the Trade Mark Registry, having regard to all the evidence, is the kind of point to which the approach to appeals summarised in Digipos applies. Mr James clearly had in mind the distinction between the different kinds of chocolate and understood the nature of the evidence before him. His decision was based on his finding on the evidence that the mark was distinctive in relation to the category "chocolate". Such a finding would justify his decision on the specification of goods but was there a proper basis for that finding in the first place?
71. Nestlé submit there was no basis for it beyond milk chocolate. Cadbury have not sought to support that finding beyond making the generalised submission I have recorded above. I have not been referred to any evidence relating to use other than on milk chocolate and I note Mr James said that Cadbury only use the colour on milk chocolate.
72. In paragraph 56 of Libertel the CJEU said:

56. The greater the number of the goods or services for which the trade mark is sought to be registered, the more excessive the exclusive right which it may confer is likely to be, and, for that very reason, the more likely is that right to come into conflict with the maintenance of a system of undistorted competition, and with the public interest in not unduly restricting the availability of colours for the other traders who market goods or services of the same type as those in respect of which registration is sought.

73. There is no doubt Mr James had this in mind overall since he cited it at paragraph 100 and recognised its importance in relation to the need to have acquired distinctive character for each and every item covered by the list of goods at paragraph 101. However as I read the passages of the Decision at the points Mr James finds that the colour purple is distinctive of Cadbury for chocolate (paragraphs 103-112 and 115), he is not there focussed on the distinction between milk chocolate and other kinds of chocolate. The focus is on other matters. Nowhere in making the finding that the colour is distinctive for chocolate in bar and tablet form or chocolate per se does Mr James consider whether the material before him, based as it is purely on the use of the colour on milk chocolate, might support a wider finding.
74. Cadbury submitted that Mr James is vastly experienced in trade mark matters. If I may say so, I entirely agree. Once he had made the finding that purple was distinctive of Cadbury for chocolate, I think the Decision at paragraph 128 was unimpeachable but although Cadbury referred in general to the vast amount of evidence before Mr James, they have not drawn my attention to any of it which would support a finding of distinctiveness wider than for milk chocolate. The question of whether there is a justification for a finding wider than milk chocolate was not considered in the Decision and I will consider it briefly now.
75. The evidence clearly supports a finding that purple is distinctive of Cadbury for milk chocolate per se. That disposes of the suggestion that this part of the specification should be limited to chocolate buttons. It should not.
76. It is also true (decision paragraph 128) that all kinds of chocolate share the same basic properties. However just because they share the same basic properties does not mean they are the same. Plain and white chocolate are clearly different enough from milk chocolate for Cadbury to draw a distinction between them in their branding. After all Cadbury do not use the colour purple as a trade mark on the other kinds of chocolate. Particularly bearing in mind the public interest in not unduly restricting the availability of colours for the other traders who market goods or services of the same type (see Libertel paragraph 56 above) in my judgment it would not be right to say that the colour purple is distinctive of chocolate generally without specific evidence dealing with other kinds of chocolate such as plain or white chocolate.
77. Accordingly the finding of distinctiveness was too wide and on that basis the specification is too broad. I will allow the appeal in this respect.

Chocolate for eating

78. Nestlé submits that Mr James should not have allowed the mark for the category "chocolate for eating". The first point taken against the category is a lack of comprehensibility contrary to Sieckmann. The term is not used in the trade or by the public and is too vague and uncertain. The second point is that the amendment of the classification to include "chocolate for eating" amounts to an impermissible extension of the coverage of goods which is precluded by s39 of the Act. Third the decision is said to be illogical because he had disallowed chocolate confectionery and chocolate assortments, which are both chocolate for eating.
79. I can deal with these points briefly. On the first point there is no doubt that the specification of goods or services must be sufficiently clear and precise as the subject matter allows to enable the authorities and third parties to determine accurately the scope of protection conferred by the trade mark (Sieckmann). Mr Malynicz also referred to the opinion of Advocate General Bot (paragraph 79) in the IP Translator case (C-307/10, Chartered Institute of Patent Attorneys v Registrar of Trade Marks). When Mr Malynicz made his submission the CJEU had not given judgment. That came after the hearing on 12^th June 2012 and paragraphs 46-49 of the judgment are to the same effect.
80. Mr James dealt with the same submission about "chocolate for eating" at paragraph 130 of the Decision. The point was that he had found the colour distinctive of Cadbury for chocolate per se but, because the specification could not be widened, it would not be permissible to use the word "chocolate" alone since that included cooking chocolate. Hence "chocolate for eating" was chosen. Nestlé argued it was vague and unfamiliar. Mr James said:

130. Everyone knows what chocolate is, and we are all familiar with the meaning of the word 'eating'. Further the term 'cooking chocolate' is well understood. I cannot therefore understand why the meaning of 'eating chocolate' should be unclear, even if it is unfamiliar. I therefore reject Nestlé's […] point.

81. I agree with Mr James.
82. The second point is that the amendment has extended the scope of the specification. This was not put to Mr James below although I have no doubt at all that he would not have allowed such an amendment if he thought it would widen the scope. Cadbury's answer to Nestlé's point is that "chocolate for eating" is a subset of "chocolate confectionery", which was in the original specification and so there is no broadening. The reason for the amendment is explained above in paragraph 10(iii). I agree with Cadbury.
83. The third point really falls with the second. The distinction Mr James drew in his decision was between products which were essentially just chocolate itself and other products consisting of a blend of chocolate with other things (see paragraphs 113 (on assortments) as well as 115 and 116 (on confectionery)). Cadbury won on chocolate itself but lost on blends. "Chocolate for eating" refers to products which are essentially just chocolate itself and not to the blends. The result is not illogical.

Conclusion

84. The appeal is dismissed save in relation to the point on milk chocolate as opposed to chocolate generally. The specification of goods will need to be amended. I will hear the parties on the point but my preliminary view is that the specification should read:

Milk chocolate in bar and tablet form; milk chocolate for eating; drinking chocolate; preparations for making drinking chocolate.

Postscript

85. When I circulated this judgment in draft two points arose. It is convenient to deal with them here and now.
86. First Nestlé submitted that paragraph 12 and paragraph 45 of the judgment do not accurately reflect Nestlé's case on distinctiveness. Mr Malynicz referred to paragraph 97 of his skeleton argument. It provides as follows:

"97. The appellant's primary case is set out above. If it is right about its primary case on section 3(1)(a) then the mark cannot be said to have acquired distinctive character under the proviso to section 3(1)(b) of the Act. The appellant does not run any discrete arguments under s3(1)(b). It confines the appeal to asserting that it is not possible for a mark (a) which is not a sign or (b) whose graphical representation does not satisfy the Sieckmann criteria, to acquire distinctive character through use."

87. Thus Nestlé's argument was that the mark applied for is not a sign capable of being represented graphically (those are the issues I addressed at length). Nestlé advanced no discrete appeal grounds on the question of acquired distinctiveness (in other words Nestlé did not challenge the important findings on the evidence subject to the issue on the category of goods). However Nestlé did submit that it was not possible for a mark that is not a sign or not properly graphically represented to acquire distinctiveness. That point is not recorded in paragraphs 12 or 45 of the judgment but it is in paragraph 97 of the skeleton, albeit that Nestlé made clear there that it did not run any discrete arguments under s3(1)(b). Thus I am happy to record that Nestlé did indeed make the point but no argument at all was advanced that it added anything to the arguments about being a sign and being represented graphically.
88. Second, in response to my preliminary view on the specification of goods at paragraph 84 above, the parties made submissions in writing.
89. Cadbury submitted that the wording proposed was appropriate in the light of the substance of the judgment. The wording proposed was compliant with the legal requirements (Sieckmann and IP Translator). Cadbury also submitted that it would not be appropriate for different or other wording to be used, all the more so when no such wording was proposed before Mr James or on Appeal.
90. Nestlé submitted that in the circumstances the appropriate wording is as follows (with the insertion marked in bold and underlined):

Milk chocolate in bar and tablet form; milk chocolate for eating; none being chocolate assortments or other products consisting of a blend of chocolate with other ingredients; drinking chocolate; preparations for making drinking chocolate.

91. Nestlé submitted this wording reflects the judgment at paragraph 83 and avoids the situation of someone consulting the register and, without seeing the Decision of Mr James or this judgment, thinking that "chocolate for eating" may cover assortments and blends. Therefore in the interests of legal certainty, following IP Translator, Nestlé invited me to consider the amended wording it proposed.
92. I prefer Cadbury's submission for the reasons given. Moreover Nestlé's proposal is in truth an attempt to re-argue the "chocolate for eating" point. But I rejected the appeal on "chocolate for eating". The reason for describing the proposed wording as a preliminary view was because it arose from allowing the appeal on the milk chocolate/plain chocolate point and I had not heard argument from the parties as to how to put that conclusion into effect. I find that the specification should be as set out in paragraph 84 of the judgment and I will make an order to that effect.