Fromageries Bel SA v J Sainsbury Plc

This is an appeal from the decision dated 13 February 2019 of Allan James sitting as a

Michael Edenborough QC and Tim Bamford appeared for the FBSA, Simon Malynicz QC for the Respondent (‘Sainsbury’s’).

The Trade Mark is registered in class 29 in relation to cheese. It was sometimes identified in the evidence by reference to FBSA’s Mini Babybel cheese. This is the pictorial representation:

The trade mark is identified to be a three dimensional mark with the following description:

“The mark is limited to the colour red. The mark consists of a three dimensional shape and is limited to the dimensions shown above.”

Before the Hearing Officer it was argued that the Trade Mark does not satisfy either s.3(1)(a) or s.3(2) of the Trade Marks Act 1994 (‘the 1994 Act’) and should therefore be declared invalid pursuant to s.47(1).

The argument under s.3(2) was dismissed. The argument under s.3(1)(a) was advanced on four grounds, of which one succeeded. It was that ‘the colour red’ in the description did not provide sufficient clarity and precision. A particular hue of red should have been specified. FBSA now appeals that finding.

Sections 1(1) and 3(1)(a) of the 1994 Act provide:

“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.

…

3.

(1) The following shall not be registered –

(a)

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

…”

In the Decision the Sieckmann criteria were discussed. These are taken from the judgment of the Court of Justice in Sieckmann v Deutsches Patent- und Markenamt (C273/00) EU:C:2002:748; [2003] Ch 487 in which the law on the graphic representation of a trade mark was explained. The main point in issue in Sieckmann was whether a trade mark may consist of a sign which cannot be perceived visually. In the course of saying that it could, the Court set out criteria, summarised in paragraph 55, which have since been applied in relation to all signs. The criteria are that the sign:

“… can be represented graphically, particularly by means of images, lines or characters, and that the representation is clear, precise, self-contained, easily accessible, intelligible, durable and objective.”

By way of a preliminary step, before considering the various grounds on which the

Dealing first with the objections to validity under subsections 3(2)(a), (b) and (c), the Hearing Officer said that his finding that the colour red is an essential characteristic of the Trade Mark was enough by itself to reach a finding that the application for invalidation based on grounds under s.3(2) failed, although he went on to consider other arguments.

Turning to s.3(1)(a) the Hearing Officer addressed the question of how many reds are shown in the pictorial representation of the Trade Mark:

“[75] … In my view, there is room for argument as to whether the part of the protrusions shown in a fuchsia-like colour fall within the description ‘red’. However, for present purposes I accept that it does. I therefore reject [Sainsburys’] claim that the presence of the parts of the mark in fuchsia and white creates a mismatch between the pictorial representation of the mark and the description, which renders the mark unintelligible.”

Having made that finding, the Hearing Officer next considered the question whether the colour red is defined with sufficient precision and clarity. He referred to the judgment of the Court of Justice on colour per se marks in Libertel Groep BV v BeneluxMerkenbureau (C-104/01) EU:C:2003:244; [2003] ECR I-3822:

“[78] The Libertel judgment has been generally understood as meaning that neither a sample of the colour(s), nor merely descriptions of colour in words, such as ‘red’, are sufficiently clear, precise and/or durable to satisfy the requirement for a graphical representation of colour marks. Instead, the practice of using internationally recognised colour identification codes has become the usual way of representing such marks.”

The Hearing Officer accepted that it was not invariably necessary to define precisely the colour or colours shown in marks. He continued:

“[81] However, in my view the Libertel requirements apply to every trade mark in which colour is an essential characteristic of the mark. If it were otherwise, the task of deciding whether two such marks were identical, or even similar, could be made impossible. And this would lead to the kind of uncertainty that the Libertel requirements are intended to avoid. [original underlining]

[82] It follows from my finding at paragraph 40 above that I consider the colour red to be an essential characteristic of the contested mark. The question whether the colour red is defined with sufficient precision is therefore of decisive significance.”

The Hearing Officer referred to UK decisions in which specificity as to colour had been in issue: Robert McBride Ltd’s Trade Mark Application [2003] RPC 19 and Calor Gas (Northern Ireland) Ltd (unrep.) (BL O/340/06) and said:

“[84] It therefore seems clear that in the case of a mark in which colour is important, such as this one, defining the colour with the broad description ‘red’ is insufficiently precise to satisfy the requirement for a graphical representation of the trade mark.

[85] I do not accept that this deficiency is cured by the presence of the cherry and fuchsia-like colours shown in the pictorial representation of the trade mark. I find the colours shown in the picture of the mark do not limit the mark to the shades of red shown. This is because (i) the description of the mark does not say so, and (ii) there is no rule that the pictorial representation of the mark takes precedence over the description so that it would be appropriate to infer that the claim to the colour ‘red’ is limited to the shades of red shown in the picture.”

With regard to his point (ii), the Hearing Officer relied on Glaxo Wellcome UK Ltd v Sandoz Ltd [2017] EWCA Civ 335; [2017] FSR 33, drawing an analogy with this case and thereby the principle that the picture in a trade mark specification could not be relied on to limit the meaning conveyed by the description.

The Hearing Officer then considered three alternatives: the primary indication of the subject matter of the trade mark should be taken (a) from the written description, (b) the picture or (c) neither.

If primary guidance comes from the written description, ‘the colour red’ does not satisfy the Libertel requirements. If it is drawn from the picture, the absence of colour identification codes means that the colours are not represented in an objective and durable manner:

“[89] … This is because the pictorial representation of the colours on the register is only as accurate as the process used to capture the colours when they were scanned from the application into the electronic register. The original representation is liable to change over time in the same way as a physical sample of the goods.”

The final alternative was that neither the pictorial representation nor the written description of the Trade Mark should be considered as secondary to the other. This would lead to ambiguity. Some would regard the mark as covering any sort of red, others only cherry red and fuchsia as shown in the picture. The Hearing Officer referred again to Glaxo, this time noting the finding of the Court of Appeal that leaving the public in a state of confusion in that way renders a mark invalid.

For those reasons the Hearing Officer found that the Trade Mark did not satisfy s.3(1)(a) of the Act and that it should be declared invalid.

The grounds of appeal as set out in the Appellant’s Notice were:

FBSA sought to adduce new evidence in support of these grounds. The first was to show that the colour of the product in the registration from which the Trade Mark takes priority, i.e. French registration no. 95/588423 of 18 September 1995 (‘the French Registration) was Pantone 193C. The second was a printout of trade mark registrations from the UKIPO database intended to show that if the Hearing Officer was right, large numbers of trade marks must be invalid. The admissibility of this new evidence on appeal was challenged by Sainsbury’s.

There was a Respondent’s Notice. This was to revive the argument, rejected at paragraph 89 of the Decision, that the colour fuchsia is not red and that the presence of

In Consolidated Developments Ltd v Cooper [2018] EWHC 1727 (Ch); [2019] FSR 2, Henry Carr J reviewed the authorities on the admission of new evidence in trade mark appeals, in particular Hunt-Wesson Inc’s Trade Mark Application [1996] RPC 234, Club Europe Trade Mark [2000] RPC 329, Wunderkind Trade Mark [200] RPC 45, EI Dupont de Nemours & Company S.T. Dupont [2003] EWCA Civ 1368; [2004] FSR 15 and Gerry Webber International AG v Guccio Gucci SpA [2015] RPC 9. He also referred to the well-known Ladd v Marshall factors (see Ladd v Marshall [1954] 1 WLR 1489, at 1491).

In Omega Engineering Inc v Omega SA [2010] EWHC 1211 (Ch); [2010] FSR 26, there were two categories of evidence. The first was from an official in the trade marks registry that goods defined in a written agreement in issue were properly classified in Class 9. The second was evidence that one of the parties owned certain trade mark registrations at the date of the agreement. Arnold J noted that by s.34(2) of the 1994 Act any decision as to classification is a matter for the Registrar, whose decision is final. He observed that both categories of evidence were a matter of public record and it was therefore debatable whether permission was needed to rely on them in an appeal from the Registry. In the event there was no opposition to the evidence being admitted.

In British Sugar plc v James Robertson & Sons Ltd [1996] RPC 281. At p.305 Jacob J said:

“Both sides invited me to have regard to the state of the register. Some traders have registered marks consisting of or incorporating the word ‘Treat’. I do not think this assists the factual inquiry one way or the other, save perhaps to confirm that this is the sort of word in which traders would like a monopoly. In particular the state of the register does not tell you what is actually happening out in the market and in any event one has no idea what the circumstances were which led the registrar to put the marks concerned on the register. It has long been held under the old Act that comparison with other marks on the register is in principle irrelevant when considering a particular mark tendered for registration, see e.g. MADAME Trade Mark and the same must be true under the 1994 Act. I disregard the state of the register evidence.”

I do not think that permission is required for FBSA to refer to the French priority registration and what it contains. That registration is not only a matter of public record, it is tied up with the third ground of appeal and must be considered in the context of that ground, which I consider below.

The position is different in relation to the many trade mark registrations which FBSA wishes to rely on, essentially in support of a ‘floodgates’ argument. Upholding the Decision would mean, Mr Edenborough warned, that vast swathes of marks on the UK Registry database would suddenly become vulnerable to a declaration of invalidity, including all of those in the new evidence.

By agreement I looked at the collection of registrations produced by Mr Edenborough de bene esse. I do not accept that upholding the Decision would have the dramatic effect claimed for reasons I discuss below. I cannot, of course, say that there could be no possible consequence to any other trade mark currently on the UK register. In my view, neither the first nor second of Ladd v Marshall factors is satisfied. Mr Edenborough did not point to any of the Hunt-Wesson factors (to the extent that they don’t overlap those of Ladd v Marshall) as a matter which should lead me to conclude that this evidence must be admitted.

I would have refused to admit the UK registrations in evidence, but they were fully debated in argument in relation to one of the key points. In fact, it is helpful to refer to two of them in the course of explaining the reasons for my finding on that point. It seems to me not appropriate to discuss two of the registrations and at the same time to exclude the registrations from the evidence, or alternatively all but two of them. So this evidence is admitted along with the French priority registration. Their admission is without prejudice to my respectful agreement with the view expressed by Jacob J in British Sugar that in principle other marks on the register are irrelevant when considering the validity of a particular mark. That remains true in the present case even though two of the registrations are used by me to illustrate an issue.

The first ground of appeal was not pursued as stated in the Appellant’s Notice. Unsurprisingly, since Sieckmann itself was about an olfactory mark (the odour created by methyl cinnamate, reported to be balsamically fruity, with a slight hint of cinnamon). In his skeleton argument Mr Edenborough accepted that the Sieckmann criteria apply to all trade marks.

His case, as adapted, was that the way in which the criteria should be applied differs, depending on the type of mark. The Sieckmann criteria require more than just precision; among other things the graphic representation must be self-contained and accessible. For a colour per se mark, it is necessary to be precise about the hue (Footnote: ¹) because it is the only characteristic of the mark. This means that the other requirements that the graphic representation must be self-contained and accessible, are necessarily compromised by requiring the reader of the specification to look up a specified Pantone number, for instance. For marks which include colour but which are not colour per se marks, precision as to the hue is not necessary. Therefore other Sieckmann requirements should not be compromised. No Pantone number or other means of achieving precision is needed and the reader of the specification should not be put to the trouble of finding it.

Mr Edenborough continued by saying that there are two types of colour mark: those where the colour is the only essential characteristic of the mark – colour per se marks

Mr Edenborough then pointed to the collection of trade marks on the UK Trade Marks Registry assembled by FBSA, all of which feature colour, none of which identify the hue or hues used precisely and all of which, he said, would become vulnerable to a declaration of invalidity if the Hearing Officer was right.

Mr Malynicz referred to the two decisions relied on by the Hearing Officer given by Geoffrey Hobbs QC sitting as an Appointed Person. In Calor Gas (Northern Ireland) Ltd’s Trade Mark Application SRIS O-340-06 (unrep.), the applicant had applied to register a series of two signs which were graphically represented in the form of a gas cylinder of the type used to fuel portable gas heaters. The representation was in monochrome and the description was as follows:

“The Trade Mark consists of the colour yellow applied to the outer surface of the cylinder within which gas is contained.”

Mr Hobbs said:

“The Sieckmann criteria are clearly not satisfied by a graphic representation which lacks precision as to the particular colour or colours in which the identity of a colour specific sign resides. That is especially true when the lack of precision is attributable to the inclusion of general designations such as ‘red’, ‘green’ or ‘blue’ in the graphic representation of the sign. Such designations are, in the absence of further clarification, apt to render the representation unacceptable for ambiguity as noted in Robert McBride Ltd's Trade Mark Application [2003] RPC 19, page 343 at paragraph 9:

‘... The identity of a colour resides in the hue that it presents to the eye of the observer. A verbal description which fails to identify the hue of the colour it seeks to define is a recipe for uncertainty: it would leave room for differing perceptions of different individuals to be equally applicable benchmarks by which to judge whether a particular hue falls within the relevant wording, with each individual also being liable to regard the wording of the description as applicable to a multiplicity of different colour shades, c.f. Orange Personal Communications Ltd's Application.’

In order to avoid that problem, the graphic representation should contain as much information as people who consult the register would realistically need to know in order to determine how closely a given colour or colour scheme matches the one which the graphic representation is intended to identify. So as to satisfy the Sieckmann requirement for durability, the graphic representation should also be recorded with the degree of permanence envisaged by the ECJ in

22 paragraphs 31 to 38 of its judgment in Case C-104/01 Libertel [2003] ECR 1-3793:”

Mr Malynicz argued that the present case was analogous to Calor Gas, raising the same issues regarding precision and permanence. He submitted that there was an important difference between almost all the registrations Mr Edenborough had referred to in the new evidence and the contested Trade Mark. The distinction was that in the present case the Hearing Officer correctly found the colour red to be one of the essential characteristics of the mark and had correctly concluded that therefore the requirements of precision and permanence identified in Calor Gas applied to this case.

In reply Mr Edenborough submitted the passage from Calor Gas quoted above, relied on by both the Hearing Officer and Mr Malynicz, was inconsistent with what Mr Hobbs had said in Robert McBride, notwithstanding the reference in Calor Gas to the latter case. Mr Edenborough drew attention to paragraph 10 of Robert McBride:

“[10] The registrar's published statements of practice have accordingly emphasised that an application for registration in colour should be based upon a graphic representation of the relevant trade mark filed in the relevant colour(s) or filed in black-and-white with the relevant colour(s) precisely defined by reference to a Pantone or other widely known and readily available colour standard: see pp.19 and 20 of Chapter 6 of the Trade Marks Registry Work Manual (June 1996); Practice Amendment Circular 9/97 (April 1997) and Practice Amendment Circular 2/00 (February 2000).”

I can summarise the principal borderline between the parties’ arguments in this way. Neither Mr Edenborough nor Mr Malynicz challenged the finding by the Hearing Officer that the colour red is an essential characteristic of the Trade Mark. They differed on whether that finding was correctly applied. Mr Edenborough’s position was that because colour was one of several essential characteristics, there was no requirement that its hue be specified in the description. Mr Malynicz supported the Hearing Officer’s approach: once he had found that the colour red was an essential characteristic, it necessarily followed that precision as to hue was required.

Before the Hearing Officer turned to s.3(1)(a) of the 1994 Act, he considered first the objections under s.3(2). There were three, one under each of subsections 3(2)(a), (b) and (c). At the relevant time that subsection provided:

“(2)

A sign shall not be registered as a trade mark if it consists exclusively of –

(a)

the shape which results from the nature of the goods themselves,

(b)

the shape of goods which is necessary to obtain a technical

result, or

(c)

the shape which gives substantial value to the goods.”

The concept of the ‘essential characteristics’ of a mark was first developed by the Court of Justice in relation to the equivalent to s.3(2)(b) of the 1994 Act found in Directive 89/104, i.e. art.3(1)(e), second indent. In Regulation (EC) 40/94 and later Regulation (EC) 207/2009, both (like Directive 89/104) formerly in force, the equivalent provision

In Koninklijke Philips Electronics NV v Remington Consumer Products Ltd (C-299/99) EU:C:2003:377; [2003] Ch 159, the Court of Justice considered the validity of a threedimensional mark by reference to the second indent of art.3(1)(e) of Directive 89/104. The Court set out the policy behind art.3(1)(e):

“[78] The rationale of the grounds for refusal of registration laid down in Article 3(1)(e) of the Directive is to prevent trade mark protection from granting its proprietor a monopoly on technical solutions or functional characteristics of a product which a user is likely to seek in the products of competitors. Article 3(1)(e) is thus intended to prevent the protection conferred by the trade mark right from being extended, beyond signs which serve to distinguish a product or service from those offered by competitors, so as to form an obstacle preventing competitors from freely offering for sale products incorporating such technical solutions or functional characteristics in competition with the proprietor of the trade mark.

[79]

As regards, in particular, signs consisting exclusively of the shape of the product necessary to obtain a technical result, listed in Article 3(1)(e), second indent, of the Directive, that provision is intended to preclude the registration of shapes whose essential characteristics perform a technical function, with the result that the exclusivity inherent in the trade mark right would limit the possibility of competitors supplying a product incorporating such a function or at least limit their freedom of choice in regard to the technical solution they wish to adopt in order to incorporate such a function in their product.

[80]

As Article 3(1)(e) of the Directive pursues an aim which is in the public interest, namely that a shape whose essential characteristics perform a technical function and were chosen to fulfil that function may be freely used by all, that provision prevents such signs and indications from being reserved to one undertaking alone because they have been registered as trade marks (see, to that effect, Windsurfing Chiemsee, paragraph 25).”

The Court ruled (at [84] and in the fourth paragraph of the final ruling):

“[84] … [Directive 89/104] must be interpreted to mean that a sign consisting exclusively of the shape of a product is unregisterable by virtue thereof if it is established that the essential functional features of that shape are attributable only to the technical result. …”

In its judgment the Court referred interchangeably to the ‘essential functional features’, the ‘essential functional characteristics’ and the ‘essential characteristics’ of the shape of a product.

In Lego Juris A/S v OHIM (Case 48/09 P) EU:C:2010:516; [2010] ETMR 63, the Court of Justice took the concept forward and explained in some detail how the assessment

The Court also considered the relevance of the perception of the average consumer:

“[76] The presumed perception of the sign by the average consumer is not a decisive element when applying the ground for refusal under art.7(1)(e)(ii) of Regulation 40/94, but, at most, may be a relevant criterion of assessment for the competent authority when it identifies the essential characteristics of the sign.”

Since then, in any assessment of validity of a three-dimensional mark under art.7(1)(e)(ii) of Regulation 40/94 (or any equivalent provision in the Trade Mark Directive, the Trade Mark Regulation or the version the 1994 Act in force at the relevant time) the tribunal has taken a necessary preliminary step, being the identification of the essential characteristics of the contested sign.

The same approach has been taken to two-dimensional marks, see Pi Design v Yoshido Metal Industry (C-337/12 P to C-340/12P) EU:C:2014:129 at [55]).

Hauck GmbH & Co KG v Stokke A/S (Case C-205/13) EU:C:2014:2233 was concerned with the first and third indents of art.3(1)(e) of Directive 89/104 (equivalent to art.7(1)(e)(i) and (iii) of Regulation 40/94 and s.3(2)(a) and (c) of the 1994 Act then in force, quoted above). The CJEU held (at [20]) that the ground for refusal of registration under the first and third indents of art.3(1)(e) pursued the same objective as that of the second indent. Consequently, in order to apply the first and third indents correctly, it was necessary to conduct the preliminary step of identifying the essential characteristics of the sign.

I am not aware of any judgment in which the CJEU or any other court has ruled that identifying the essential characteristics of a contested sign is a necessary preliminary step in the assessment of any other ground of invalidity of a trade mark. That includes the version of s.3(1)(a) of the 1994 Act relevant to this appeal and its equivalents.

Returning to art.7(1)(e)(ii) of Regulation 40/94, the CJEU stated in Lego that identifying the essential characteristics of a sign requires the tribunal to decide which are the most important elements of the sign. This is largely a visual assessment which can be approached by considering the overall impression of the sign or by examining in turn each of the components of the sign, as appears from the paragraphs of the judgment quoted above.

But it is not quite that simple. There may be a feature of a sign which would plainly strike the eye of any observer but which does not qualify as an essential characteristic. In Lego the sign in respect of which registration was sought was a conspicuously red Lego brick. The Court of Justice ruled that the colour red was nonetheless not an essential characteristic of the sign:

“[51] As regards the fact that the ground for refusal covers any sign consisting ‘exclusively’ of the shape of goods which is necessary to obtain a technical result, the General Court stated, at [38] of the judgment under appeal, that that condition is fulfilled when all the essential characteristics of a shape perform a technical function, the presence of non-essential characteristics with no technical function being irrelevant in that context.

[52] That interpretation is consistent with [79] of Philips. Moreover, it reflects the idea underlying that judgment, as set out by Advocate General RuízJarabo Colomer at point 28 of his Opinion in that case and also at point 72 of his Opinion in Case C-363/99 Koninklijke KPN Nederland [2004] ECR I-1619, that is to say, that the presence of one or more minor arbitrary elements in a three-dimensional sign, all of whose essential characteristics are dictated by the technical solution to which that sign gives effect, does not alter the conclusion that the sign consists exclusively of the shape of goods which is necessary to obtain a technical result.”

Paragraph 79 of Philips is quoted above. This is paragraph 28 of the Opinion of Advocate General Ruíz-Jarabo Colomer in the same case, also cited by the Court, (with paragraph 27 to put it into context):

“[27] The national court seeks to ascertain, in particular, whether a merely functional shape is caught by the exclusion provided for in subparagraph (e) even when it can be shown that the same technical result can be achieved by other different shapes.

[28] ‘Merely functional’ is to be understood - as suggested by the national court - as any function whose essential features are attributable to the achievement of a technical result. Use of the phrase ‘essential features’ means that a shape containing an arbitrary element which, from a functional point of view, is minor, such as its colour, does not escape the prohibition.”

In paragraph 72 of his Opinion in KPN Nederland Advocate General Ruíz-Jarabo Colomer repeated what he had said in Philips, this time italicising the words ‘which, from a functional point of view, is minor’.

There may be something circular about this. A conspicuous feature of a sign, such as its colour, can be excluded from qualifying as one of the sign’s essential characteristics and therefore excluded from an assessment under art.7(1)(e)(ii) as to whether any of those characteristics is not functional, because the colour is not functional. Had the description in Lego identified a Pantone number for the red colour, the colour may, it seems, still have been dismissed as arbitrary because it is not a functional feature.

There cannot be an exact transfer of the approach to identifying the essential characteristics of a sign from used in Lego to all other contexts, otherwise colour could never qualify as an essential characteristic of a sign unless it has a function. To that extent, the test for an essential characteristic will not be identical in the context of art.7(1)(e)(i) and (iii) and might be different again if it can be applied outside art.7(1)(e) (and equivalents). If the exception to a purely visual assessment has analogies outside the context of art.7(1)(e)(ii) and equivalents, it is not clear how the analogies would work.

There is a further complication. If determining whether precision as to hue is required for a colour in a sign depends on deciding whether the colour is an essential characteristic of the sign, and to the extent that this requires a visual assessment to identify the most important elements of the sign, the visual criterion for that assessment is not self-evident. It could be that any element of the sign which makes more than an insignificant impression on the eye is one of the most important elements. If so, the colour in almost all trade marks is an essential characteristic requiring precision as to hue. If not, it is difficult to know where the line should be drawn.

I am not sure that tying the issue in hand – whether the hue of a colour featured in a mark must be defined precisely – to the essential characteristics of the sign within the meaning accorded to that term by art.7(1)(e) of Regulation 40/94 is either necessary or helpful.

The issue in hand depends on s.3(1)(a) of the 1994 Act, which refers back to s.1(1). I return to Libertel. The Court of Justice discussed the requirements of the colour in a colour per se mark by reference to art.2 of Directive 89/104, equivalent to s.1(1) of the 1994 Act:

“[21] The questions referred, which concern Article 3 of the Directive, relate to whether, and if so in what circumstances, a colour per se, not spatially defined, is capable of possessing distinctive character for certain goods or services.

[22]

In order to consider those questions it is necessary as a preliminary matter to determine whether a colour per se is capable of constituting a trade mark for the purposes of Article 2 of the Directive.

[23]

To that end, the colour must satisfy three conditions. First, it must be a sign. Secondly, that sign must be capable of graphic representation. Thirdly, the sign must be capable of distinguishing the goods or services of one undertaking from those of other undertakings.”

The Court went on to state that a colour per se mark must satisfy the Sieckmann criteria:

“[28] Furthermore, as the Court has held, a graphic representation within the meaning of Article 2 of the Directive must enable the sign to be represented visually, particularly by means of images, lines or characters, so that it can be precisely identified (Case C-273/00 Sieckmann [2002] ECR I-11737, paragraph 46).

[29] In order to fulfil its function, the graphic representation within the meaning of Article 2 of the Directive must be clear, precise, self-contained, easily accessible, intelligible, durable and objective (Sieckmann, paragraphs 47 to 55).”

The Court then held that filing a sample of the colour does not by itself satisfy the criterion of durability and therefore would not constitute a graphic representation within the meaning of art.2 of Directive 89/104. The Court continued:

“[34] On the other hand, a verbal description of a colour, in so far as it is composed of words which themselves are made up of letters, does constitute a graphic representation of the colour (see Sieckmann, paragraph 70).

[35]

A description in words of the colour will not necessarily satisfy the conditions set out in paragraphs 28 and 29 of this judgment in every instance. That is a question which must be evaluated in the light of the circumstances of each individual case.

[36]

A sample of a colour, combined with a description in words of that colour, may therefore constitute a graphic representation within the meaning of Article 2 of the Directive, provided that the description is clear, precise, selfcontained, easily accessible, intelligible, and objective.

[37]

For the same reasons as those set out at paragraph 34 of this judgment, the designation of a colour using an internationally recognised identification code may be considered to constitute a graphic representation. Such codes are deemed to be precise and stable.

[38]

Where a sample of a colour, together with a description in words, does not satisfy the conditions laid down in Article 2 of the Directive in order for it to constitute a graphic representation because, inter alia, it lacks precision or durability, that deficiency may, depending on the facts, be remedied by adding a colour designation from an internationally recognised identification code.

[39]

As to the question whether a colour per se is capable of distinguishing the goods or services of one undertaking from those of other undertakings, within the meaning of Article 2 of the Directive, it is necessary to determine whether or not colours per se are capable of conveying specific information, in particular as to the origin of a product or service.

[40]

In that connection, it must be borne in mind that, whilst colours are capable of conveying certain associations of ideas, and of arousing feelings, they possess little inherent capacity for communicating specific information, especially since they are commonly and widely used, because of their appeal, in order to advertise and market goods or services, without any specific message.

[41]

However, that factual finding would not justify the conclusion that colours per se cannot, as a matter of principle, be considered to be capable of distinguishing the goods or services of one undertaking from those of other undertakings. The possibility that a colour per se may in some circumstances serve as a badge of origin of the goods or services of an undertaking cannot be ruled out. It must therefore be accepted that colours per se may be capable of distinguishing the goods or services of one undertaking from those of other undertakings, within the meaning of Article 2 of the Directive.

[42]

It follows from the foregoing that, where the conditions described above apply, a colour per se is capable of constituting a trade mark within the meaning of Article 2 of the Directive.”

In paragraphs 34 to 38 the Court indicated that provided there are words of description of the colour, this may constitute a graphic representation. Whether or not the Sieckmann criteria are satisfied will depend on the facts. Use of a colour identification code (such as a Pantone number) will assist. In paragraphs 39 to 41, the Court considered the third requirement of a trade mark in art.2: that it must be capable of distinguishing the goods or services of one undertaking from those of other undertakings in particular as to the origin of the product or service. This also will depend on the facts.

There is a connection between the Sieckmann criteria and the requirement that the mark is capable of distinguishing the goods or services of one undertaking from those of another (hereafter ‘capable of distinguishing’). Use of a colour identification code, thereby pinning down the colour to a particular and easily identified hue, is liable to assist not only in satisfying the Sieckmann criteria but also in ensuring that the mark is capable of distinguishing. Since Libertel it has largely been assumed that a colour identification code will be necessary for a colour per se mark.

Turning to marks containing colour which are not colour per se marks, it is of course the entire mark, including non-colour elements, which must be capable of distinguishing. However, the colour element may play a part in ensuring that it is and that in turn may depend on the colour being of a particular hue.

Existing UK registrations on which FBSA sought to rely by way of additional evidence provide useful examples. It is enough to mention two of them although others illustrate the same point. One consists of the words ‘Coca Cola’, in the distinctive script long since used to advertise that drink, against a background red, silver, white and yellow. The description states ‘The applicant claims the colours red, silver, white and yellow as an element of the mark’:

The other example takes the form of the word TESCO in red capitals and beneath each letter a flat diamond shape in blue. The description states ‘The applicant claims the colours red and blue as an element of the mark’:

The point about these examples is that the precise hue is unlikely to play a significant role in making the mark capable of distinguishing. The hues could vary without affecting the mark’s capacity to distinguish.

That will not always be the case. It seems to me that where a mark contains colour but is not a colour per se mark, the need for precision as to hue will depend on the extent to which other elements of the mark serve to make the mark capable of distinguishing. More exactly, it will depend on the extent to which the colour of the relevant feature of the mark contributes to making the mark capable of distinguishing and whether it is likely that only a particular hue will confer on the mark that capacity to distinguish. It will always be a question of fact and degree.

This is consistent not only with Libertel but also with decisions I have mentioned earlier, even though the question of precision as to hue was not approached in those decisions in quite this way. In Calor Gas there was a finding that failing to identify the particular hue of yellow did not satisfy the Sieckmann criteria and was inconsistent with Libertel. The sign had two features: the colour yellow being specified for the outer surface of the cylinder and the shape of the cylinder. There was nothing obviously distinctive about the shape, nor was there said to be. The decision is therefore consistent with the view that the capacity of the mark to distinguish was likely to have depended on the particular hue of yellow on the surface of the cylinders. There was also a finding that the imprecision of the word ‘yellow’ had been compounded by coupling it with references to a colour which would often be thought of as orange, but this did not undermine the primary finding.

The decision in Robert McBride must in part be treated with caution because it predates both Sieckmann and Libertel. For that reason I can attach no weight to the contemporary practice of the Registrar on which Mr Edenborough relied, referred to in paragraph 10 of the decision. Otherwise the decision is fully consistent with requiring precision as to hue if it is likely to be required to confer on a sign the capacity to distinguish. The mark applied for was a series of two signs represented as threedimensional marks in the shape of a rectangular dishwasher tablet each divided into a top and bottom half. The top half of the first sign was stated to be yellow and the bottom half white, with the colours reversed for the second sign.

The essential characteristics of the Trade Mark in the present case were identified by the Hearing Officer in his paragraph 40, namely the shape of the product, its dimensions, the pull tags and the colour red. In fact it provides a summary of all the characteristics of the Trade Mark save one: the fuchsia colour of the pull tags fading to white at their ends.

Before the Hearing Officer there was evidence from Michelle Bamford, Technical and Regulatory Affairs Manager of Bel UK Ltd, the UK subsidiary of FBSA. She said:

“5.

The trade mark in its specific shape and size combined with the colour red is relied upon and recognised by consumers to identify Mini Babybel as originating from Fromageries Bel, …”

Although Ms Bamford provided evidence of the money spent on advertising Mini Babybel in the UK and its market share in this country, the part of her evidence I have just quoted was no more than an assertion. There had been an earlier statement on behalf of Sainsbury’s from Nicholas Walker, a partner in Lewis Silkin LLP. He explained that the Trade Mark is never used without other distinctive branding:

“10.

In addition, as far as I am aware, the purchasing public never see just the waxed cheese (in the form shown in the trade mark registration), in a retail situation. Where the cheeses are sold, they are always covered by at least two, and often three, layers of branding:

a)

the waxed cheese is covered by a branded wrapper;

b)

the wrapped cheese is then packed in a net bag with a branded swing tag; and

c)

the net bags of cheese generally appear on shop shelves within a branded box.”

The retail branding takes the form of labelling which prominently features the trade name ‘Babybel’ in a particular script:

Mr Walker’s evidence was not contradicted by FBSA. It seems to me very likely that in the marketplace customers rely on this labelling to identify FBSA’s products.

There was no market evidence of the extent to which a precise hue of red would be required to confer distinctiveness on the Trade Mark because it is never used without other indicia. Ms Bamford’s evidence shows that shape, size and colour are together the candidate elements of the Trade Mark for recognition by consumers to identify the source of the cheese, if ever the Trade Mark were to be used without other indicia. Whether they would in fact serve that purpose was speculation and it is not even clear whether, when Ms Bamford referred to colour, she had in mind a particular hue used by FBSA or alternatively any red at all.

For the reasons I have stated I do not accept, as a principle of law, that in respect of a trade mark featuring colour, the question whether the particular hue of the colour must be specified depends on whether the colour is an essential characteristic as that term has been used in relation to art.7(1)(e) of Regulation 40/94 and equivalents. I must therefore answer that question according to the alternative criterion I have identified, on the evidence available.

The question in more detail is whether the Trade Mark is capable of distinguishing the cheese of FBSA from the cheese of other undertakings on the assumption that the hue used in the Trade Mark may be any hue of red which FBSA cares to use and indeed which FBSA is free to vary from time to time. The possibilities multiply if the picture is taken to show that there are two hues of red so that the hue used for the main body of the puck will differ from that used for the pull tags.

I have reached the view on the balance of probabilities that the Trade Mark could be capable of distinguishing only if a particular hue of red used on the main body of the product is associated with FBSA’s cheese. It follows that the Trade Mark must be limited to a single hue of red. I reject the first ground of appeal.

FBSA’s alternative argument was that the only red to which the description could be referring is the hue shown in the pictorial representation. The picture and the mark must be treated as a whole. This last proposition was a reference to Kitchin LJ’s judgment in Glaxo, at [63]-[64], a passage also relied on by the Hearing Officer. Mr Edenborough submitted that where a trade mark is filed with a description referring to a colour and a picture representing the colour, invariably the description will be taken to mean the hue of the colour shown in the picture.

Mr Malynicz agreed with the Mr Edenborough’s proposition taken from Glaxo but argued that it had the opposite effect. The picture does not take precedence over the description. The description specified the colour red, but not any particular hue. The absence of any statement about hue (such as a Pantone number) was all the more striking because the description did expressly limit the shape to particular dimensions shown in the picture. Had the Trade Mark been filed with no description at all, the reader would of necessity fall back on the picture and assume that the colour of the

Mr Malynicz also supported the Hearing Officer’s reasoning that if reader did not reach this conclusion, the only viable alternative view would be that the description and the picture are inconsistent. The principle that such inconsistency results in a trade mark registration lacking the clarity and precision required by the Sieckmann criteria, first stated in Glaxo, has also been adopted by the CJEU, most recently in Red Bull GmbH v EUIPO (Case C-124/18 P) EU:C:2019:641:

“[37] Where the application is accompanied by a verbal description of the sign, that description must serve to clarify the subject matter and scope of the protection sought under trade mark law and such a description cannot be inconsistent with the graphic representation (see, to that effect, judgment of 27 March 2019, Hartwall, C-578/17, EU:C:2019:261, paragraphs 39 and 40).”

The point was more fully explained in Hartwall:

“[39] The verbal description of the sign serves to clarify the subject matter and scope of the protection sought under trade mark law (see, to that effect, judgment of 27 November 2003, Shield Mark, C-283/01,

EU:C:2003:641, paragraph 59, and, as an example, judgment of 24 June 2004, Heidelberger Bauchemie, C-49/02, EU:C:2004:384, paragraph 34).

Mr Malynicz had at one stage a point on Sieckmann durability. During the hearing he accepted that the digital filing by FBSA satisfied the criterion of durability so the point fell away.

I agree with the Hearing Officer for the reasons he gave, as further elaborated by Mr Malynicz. I reject the second ground of appeal.

The third ground was not really a ground of appeal. It was an application by FBSA pursuant to s.13 of the 1994 Act to limit the rights conferred by the Trade Mark by way of specifying that the colour red was Pantone 193C. The application was made by reference to the French national trade mark no. 95/588423 from which the Trade Mark takes priority.

The pictorial representation in the French mark is the same and the description is:

“La marque se caractérise par la couleur rouge du produit, sa forme de palet et ses dimensions.”

There is a letter dated 20 September 1996 to the Registry from FBSA’s trade mark attorneys proposing an amendment of the description, to its present form, in order to bring it in line with the French priority registration. In other words, it was at that time both FBSA’s view and its representation to the Registry that the two descriptions in effect mean the same thing. I think that FBSA’s view in September 1996 was correct. Where the UK description limits the mark to the colour red, that could only be the colour of the product since the product is the mark. The French registration does not imply that this colour must be any particular hue. Even if the French registration could in principle assist FBSA in its application under s.13, in practice it cannot.

Section 13 of the 1994 Act provides, so far as is relevant:

“13 (1) An applicant for registration of a trade mark, or the proprietor of a registered trade mark, may –

…

(b)

agree that the rights conferred by the registration shall be subject to a specified territorial or other limitation;

and where the registration of a trade mark is subject to a disclaimer or limitation, the rights conferred by section 9 (rights conferred by registered trade mark) are restricted accordingly.

(2)

Provision shall be made by rules as to the publication and entry in the register of a disclaimer or limitation.”

Section 13 was discussed in Nestlé SA’s Trade Mark Application [2004] EWCA Civ 1008 (Ch); [2005] RPC 5. Nestlé concerned an application for a three-dimensional mark in the form of Polo mint without the word ‘Polo’ on it. Mummery LJ (with whom Lord Phillips MR and Chadwick LJ agreed) said:

“[31] The sign in Nestlé's original application for registration consists of the three-dimensional shape of a sweet. The sign is only registrable if it is distinctive. The sign is not inherently distinctive. The hearing officer found that, on the evidence of use of the sign, acquired distinctiveness existed in relation to Nestlé’s mint flavoured sweets for the sign only if it is also white and of a given size. Those particular features of colour and size are necessary elements of the mark, if it is to match the degree of acquired distinctiveness established on the evidence; but those particular distinctive features are both absent from the mark, for which registration was originally sought.

[32]

A requirement that the elements of colour and size should be included in the mark in order to make it distinctive and therefore registrable, does not, in our judgment, fall within the provisions of s.13(1) that the applicant ‘may agree that the rights conferred by the registration shall be subject to a specified limitation.’ The required inclusion of the colour and size elements is not a specified limitation on ‘the rights conferred by the registration’ of the mark identified in the application. The requirements of colour and size do not limit ‘the rights’, which could have been acquired on registration without including them. The requirements relate to the inclusion of essential ingredients in the content of the mark in order to make it distinctive, thus satisfying the requirements for registration of the mark and thereby obtaining the rights conferred by it.

[33]

It is important to keep in mind that the ‘rights conferred by the registration’ – which, under s.13(1)(b), an applicant may agree shall ‘be subject to a specified territorial or other limitation’ – are rights conferred by s.9(1) and which would otherwise be infringed by use within s.10 of the Act. An applicant who agrees that the rights conferred by registration shall be subject to a limitation is agreeing, in effect, that the use of the mark outside the limitation is not to be treated as an infringement of the mark notwithstanding that such use would, otherwise, fall within s.10 of the Act. That is not at all the same thing as incorporating limitations in the description of the mark itself.”

There is thus an important distinction between a limitation which narrows the scope of acts which would infringe a trade mark and one which would affect the description of the mark itself.

In my view, the application by FBSA in the present case is analogous to that made by Nestlé and is not permissible for the same reason. A limitation to Pantone 193C would not merely limit the rights conferred by section 9 referred to in s.13 of the 1994 Act. It would go further and introduce an additional feature into the content of the Trade Mark in order to make it distinctive. It would affect the description of the mark itself.

I therefore dismiss FBSA’s application to limit the Trade Mark under s.13.

Mr Malynicz submitted that at paragraph 75 of his Decision the Hearing Officer came very close to deciding that the fuchsia colour of the pull tags is not red. He should have so decided and gone on to find that this provided an additional reason for the mark lacking clarity and precision since the description did not match what was shown in the picture.

I think a reader of the specification would conclude that ‘the colour red’ encompasses the colour of the pull tags. This is also consistent with the findings above.

The appeal is dismissed.