Mettenheimer and Another v Zonquasdrif Vineyards CC and Others (965/12) [2013] ZASCA 152; 2014 (2) SA 204 (SCA); [2014] 1 All SA 645 (SCA); 2013 BIP 302 (SCA) (19 November 2013)

THE SUPREME COURT

OF APPEAL OF SOUTH AFRICA

JUDGMENT

REPORTABLE Case No: 965/12 In the matter between:

ALEXANDER

ANATOLE THEODOR METTENHEIMER ..................................FIRST

APPELLANT

ZONQUASDRIFT ESTATES (PTY) LIMITED ...............................................SECOND

APPELLANT and

ZONQUASDRIF

VINEYARDS CC ..................................................................FIRST

RESPONDENT

REGISTRAR OF CLOSE CORPORATIONS ............................................SECOND

RESPONDENT

REGISTRAR OF TRADE

MARKS .................................................................THIRD

RESPONDENT Neutral citation: Mettenheimer v Zonquasdrif Vineyards CC (965/12) [2013] ZASCA 152 (19 November 2013) Coram: Brand, Theron, Pillay, Petse JJA et Meyer AJA Heard: 4 November 2013 Delivered: 19 November 2013 Summary: (a) Trade Marks Act 194 of 1993 – appellants’ trade mark registered in respect of wine – used by the respondent in the course of trade with reference to wine grapes – whether goods so similar as to give rise to the likelihood of deception or confusion as contemplated in s 34(1)(b) of the Act. (b) Close Corporations Act 69 of 1984 – whether respondent’s name calculated to cause damage to the appellants as contemplated in s 20(2)(b) of the Act.

ORDER On appeal from: Western Cape High Court, Cape Town (Davis J sitting as court of first instance): The appeal is dismissed with costs, including the costs of two counsel whenever employed.

JUDGMENT

BRAND JA (THERON, PILLAY, PETSE JJA et MEYER AJA concurring): [1] This appeal turns on the application of s 34(1)(b) of the Trade Marks Act 194 of 1993 and s 20(2)(b) of the Close Corporations Act 69 of 1984 prior to its amendment that came into operation in 2011. The first appellant, Mr Alexander Mettenheimer, is the registered proprietor of the trade mark Zonquasdrift in class 33 in respect of alcoholic beverages, except beer. The trade mark therefore covers wine but not wine grapes which, incidentally, falls in class 31. Mettenheimer and his wife are also the shareholders in the second appellant, a private company which is the owner of a farm called Zonquasdrift between Malmesbury and Riebeek Kasteel in the Western Cape. Until recently the second appellant was known as Almett Properties (Pty) Ltd, but on 3 May 2012 – that is about two years after the commencement of the present litigation – it formally changed its name to Zonquasdrift Estates (Pty) Ltd. [2] The first respondent is a close corporation with the registered name Zonquasdrif Vineyards CC. It conducts its farming business on a farm situated about one kilometre from the second appellant’s farm where it grows wine grapes which it sells under its registered name. The second and third respondents are the Registrar of Close Corporations and the Registrar of Trade Marks, respectively, who were joined in the proceedings

for the sake of formality and who never became active participants at any stage. Hence I propose to refer to the two appellants as ‘the appellants’ and to the first respondent as ‘the respondent’. [3] Proceedings started when the appellants brought an application against the respondent in the Western Cape High Court, Cape Town. The relief they sought was twofold. First, that the respondent be interdicted in terms of s 34(1)(b) of the Trade Marks Act from infringing its trade mark Zonquasdrift by, inter alia, selling wine grapes under that name. Secondly, for an order declaring that the respondent’s registered name is calculated to cause damage to the appellants and ordering the respondent, for that reason, to change its name in terms of s 20(2)(b) of the Close Corporations Act. The respondent not only opposed the relief sought on its merits, but also brought a counter-application for the removal of the appellants’ trade mark from the register on the basis that it exclusively serves to ‘designate the geographical area of the goods’ as envisaged by s 10(2)(b) of the Trade Marks Act. [4 ] With reference to the counter-application, the respondent made it clear from the outset that it never had any interest in selling goods covered by class 33 and that, absent the appellants’ application for an interdict, it would ordinarily have no interest in their registered mark. In essence, the counter-application thus amounted to no more than an additional

defence to the appellants’ claim. When the matter came before Davis J in the court a quo he found the appellants’

application wanting on its own merits with regard to both aspects of the relief claimed. In this light he considered it unnecessary to deal with the substance of the respondent’s counter-application. Hence he refused the appellants’

application with costs and dismissed the respondent’s counter-application, making no order as to costs in respect of the latter. The appeal against these orders is with the leave of this court. There is no cross-appeal in respect of the counter-application. In consequence, that application is no longer of any relevance, save to the extent that the appellants

seek to replace the court a quo’s order with regard to the costs of the counter-application with one that the respondent

should pay these costs. [5] The issues that arose on appeal will be better understood against the background facts that follow. I find it convenient to start with the geographical layout of the area involved. Zonquasdrift is the name of a natural crossing over the Berg River or a ‘drift’ in Dutch, which eventually became a ‘drif’ in Afrikaans. Historians tell us that the existence of the crossing has been well documented under that name since at least 1660 when the Dutch colonists first used it to cross the natural hindrance to their passage into the interior, created by the Berg River. In more modern times Zonquasdrift has also become known as the end of the first stage of the four day Berg River canoe marathon. The farm Zonquasdrift was originally granted in the early 1700s. Rather self-evidently, the crossing is on that original grant.

The farm of the appellants is that subdivision of the original grant which retained the original name. On the one side of the

Berg River the crossing is now on the appellants’ farm. On the opposite side of the river the crossing is on the farm Klein Zonquasdrift. The farm on which the respondent conducts its business is on the same side of the river as Klein Zonquasdrift. It is about 300 metres from the crossing and, as I have said, about one kilometre from the appellants’ farm. It is a subdivision of the farm Bosplaas. Until it was given the name Zonquasdrif Vineyards in 2008, it was never known by that name. [6] The appellants are not involved in wine-making. They do not have a cellar on Zonquasdrift farm nor do they employ any wine-makers. It is a wine grape producing farm that delivers its grapes in bulk to a co-operative cellar, Riebeek Cellars. The vinification process is conducted exclusively by Riebeek Cellars, which sources its wine grapes from 32 different farms in the Swartland area, being the members of the co-operative. The wine is then sold under the trade mark Riebeek Cellars. As a result of this communal system, it is generally not possible to say, with reference to a particular wine, where the grapes originally came from. The grapes are pooled by Riebeek Cellars and the wine sold under its label with the designation that it is a wine of origin Swartland under the wine of origin scheme, to which I shall presently return. [7] An exception to the general exposition conveyed thus far,

pertains to wine made by Riebeek Cellars from grapes of the chenin blanc cultivar that originates from two blocks on the appellants’ farm – about six hectares in extent – which had been registered under the wine of origin scheme as a single vineyard. Pursuant to an arrangement between the appellants and Riebeek Cellars, chenin blanc wine from this single vineyard is kept separate and bottled under the special Zonquasdrift label. This wine is then exported by the appellants for sale in Germany. It also bears the designation wine of origin Swartland. Moreover Riebeek Cellars has in 2009 bottled a chenin blanc wine from the appellants’ single vineyard as ‘Limited Release Chenin Blanc’ under the Riebeek Cellars trade mark, but with the indication on

the label that it comes ‘from a designated block of vines in a 47-year old single vineyard on the farm Zonquasdrift in the Swartland’. In short, the appellants’ business is therefore that of a wine grape producer and on a more limited scale, that of a wine exporter under its registered trade mark. What should perhaps be pointed out for the sake of completeness, is that the appellants never registered any trade mark for their wine grapes in class 31. [8] The respondent was incorporated on 4 August 2008. Its majority member and the deponent to its answering affidavit is Mr Petrus Walters. In the same way as the appellants, the respondent is also a wine grape producer. But, unlike the appellants, it has never sold wines under any name. Neither has it ever made any wine. It sells its wine grapes in bulk under its registered name, Zonquasdrif Vineyards, to a corporate wine cellar in the Worcester area. By contrast, the farm Klein Zonquasdrift, which is directly across the Berg River from the appellants’ farm, produces table grapes. Table grapes are not used to produce wine. Conversely, wine grapes are not suitable for consumption as a fruit in the way that table grapes are. In consequence, wine grapes are not sold to the general public, but only to wine cellars or wine co-operatives. As to why the respondent was given the corporate and trading name Zonquasdrif Vineyards CC, Walters

explained that he found the original name of the farm on which the respondent conducts business, Bosplaas, rather uninspiring. In consequence he decided to link its name to the historical nearby crossing, which is well-known to the public, so as to indicate the geographical location of the respondent’s farming business. [9] In delineating the issues it is perhaps useful to identify that which falls outside the ambit of this appeal. The appellants did not rely on passing off or any other form of unlawful competition to justify any of the relief sought in the court a quo. The case relied solely on an infringement of their trade mark rights. Potential confusion resulting from similar names of farms situated in close proximity to one another, for instance, is therefore of no consequence. Equally irrelevant is the respondent’s contention that the appellants had failed to establish a reputation either to its grape production or its wine. The fact that the appellants thus far have only sold their wine in Germany, is not material. What we have to look at is the ‘notional use’ that the appellants can make of their trade mark, which would obviously include the marketing of wine under that label in South Africa (see eg Plascon-Evans Paints Ltd v Van Riebeeck Paints (Pty) Ltd [1984] ZASCA 51; 1984 (3) SA 623 (A) at 641E-I). [10] The interdict application, as I have said by way of

introduction, rested squarely on s 34(1)(b) of the Trade Marks Act. In relevant part this section provides: ‘(1) The rights acquired by registration of a trade mark shall be infringed by- (a) . . . (b) the unauthorized use of a mark which is identical or similar to the trade mark registered, in the course of trade in relation to goods or services which are so similar to the goods or services in respect of which the trade mark is registered, that in such use there exists the likelihood of deception or confusion.’ [11] The respondent clearly used its impugned mark in the course of trade and it is not suggested that it had been authorised by the appellants to do so. That limits the enquiry to the issue of confusing similarity. Unlike s 34(1)(a), the provisions of s 34(1)(b) do not require that the offending mark be used in relation to goods in the class for which the trade mark had been registered.

It contemplates two elements, namely, (a) a mark identical or similar to the trade mark used in relation (b) to goods which are so similar to those for which it had been registered, that it gives rise to a likelihood of deception or confusion. As to the relationship between these two elements, I agree with the sentiments expressed by Thring J when he said in New Media Publishing (Pty) Ltd v Eating Out Web Services CC [2005] ZAWCHC 20; 2005 (5) SA 388 (C) at 394C-F: ‘There is, it seems to me, an interdependence between the two legs of the inquiry: the less the similarity between the respective goods or services of the parties, the greater will be the degree of resemblance required between their respective marks before it can be said that there is a likelihood of deception or confusion in the use of the allegedly offending mark, and vice versa. Of course, if the respective goods or services of the parties are so dissimilar to each other that there is no likelihood of deception or confusion, the use by the respondent even of a mark which is identical to the applicant's registered mark will not constitute an infringement; also, if the two marks are sufficiently dissimilar to each other no amount of similarity between the respective goods or services of the parties will suffice to bring about an infringement. . . .’ (See also Canon Kabushiki Kaisha v Metro-Goldwyn-Mayer Inc [1999] RPC 117 (ECJ) para 17.) [12] In comparing the two marks I find them virtually identical. Although the respondent’s mark contains the additional

reference to ‘Vineyards’ I regard the dominant or striking feature of the mark, in the sense that it will impact on the mind and recollection of the notional ordinary purchaser, in ‘Zonquasdrif’ (see eg Plascon-Evans Paints Ltd v Van Riebeek Paints (Pty) Ltd [1984] ZASCA 51; 1984 (3) SA 623 (A) at 641D-E). The further difference between the two, which lies in the absence of the final ‘t’ in the respondent’s mark – changing it from Dutch into Afrikaans – is in my view insignificant. In the unlikely event that it is noticed by the notional purchaser, it could be ascribed to a misspelling of the appellants’ mark. What it therefore boils down to in the end is the similarity of the goods. Can it be said that, having regard to the sameness of the two marks, the similarity between the goods in respect of which the appellants’ mark is registered (wine) and the goods in

which respondent trades (wine grapes) is such that confusion or deception is the probable result? The deception and confusion we are talking about, of course, must relate to the origin of the respective goods. So, is it likely that the notional purchaser may be confused to think that these goods have the same origin? [13] Considerations that could assist in determining this likelihood of confusion have been proposed in decided cases. Prominent

amongst these are the judgments of the Chancery Division in British Sugar Plc v James Robertson & Sons Ltd [1996] RPC 281 (Ch D) at 296-297 and of this court, albeit in a context different from s 34(1)(b) in Danco Clothing (Pty) Ltd v Nu-Care Marketing Sales and Promotions (Pty) Ltd [1991] ZASCA 121; 1991 (4) SA 850 (A) at 860-861. Included amongst the considerations proposed in these cases are: (a) the uses of the respective goods; (b) the users of the respective goods; (c) the physical nature of the goods; and (d) the respective trade channels through which the goods reach the market. It is clear that the list of proposed considerations was never intended to be exhaustive. Yet they serve to stimulate and guide the required process of reasoning in the enquiry as to the likelihood of confusion. [14] On application of these considerations to wine grapes and wine, first impressions are that the likelihood of confusion is slight indeed. The nature of the two products is entirely different. The one is a fruit – albeit inedible – and the other is an alcoholic beverage. As are their uses, their users and the trade channels through which they are marketed. Since wine grapes are not suitable for consumption as a fruit, they are not sold to the public and they are therefore not to be found in any retail outlets. Wine, on the other hand, is marketed, advertised and sold directly to the

public in supermarkets, liquor stores and other retail outlets. The prospects of Zonquasdrift wine and Zonquasdrif grapes ever being marketed or sold in close proximity can therefore safely be excluded as non-existent. [15] The appellants’ argument as to why, despite these distinct differences between the two products, they have succeeded in showing a likelihood of confusion, went along the following lines. Wine grapes constitute the raw material from which wine is made. Hence the two products are associated with one another. It is known in the industry that while some farms sell their wine grapes to wineries, others make their own wine from grapes produced on the farm or purchased from other farms. There is therefore no clear delineation between farms producing and selling grapes and those selling both wine and grapes. Although the appellants do not make their own wine and only sell their Zonquasdrift wine in Germany, they are entitled to sell their wine in South Africa as well. What must also be postulated as part of the appellants’ notional use

is that they may integrate their grape growing and wine-making activities on the farm Zonquasdrift. In these circumstances, so the appellants’ argument went, the notional member of the public may very well be confused in thinking that the appellants’ wine and the respondent’s grapes originate from the same farm. [16] In evaluating the appellants’ argument, it seems to me that it draws no distinction between the notional purchasers of grapes, on the one hand, and the notional purchasers of wine, on the other. But, in the light of the distinct differences between the two products that I have alluded to, I do not believe logical reasoning allows this conflation. For the very reason that wine grapes are suitable for wine-making only, they are exclusively sold to wine makers and wine co-operatives for that purpose. The undisputed evidence by the respondent is that wine cellars and wineries purchase grapes based on a number of factors, including the cultivar and the terroir, the vineyards and the like but not with reference to the trade names of wine or the names of farms. Furthermore, wineries and co-operative cellars make use of specialists in the wine industry to buy their wine grapes from approved suppliers who comply with their own established quality control standards. The chances of these buyers confusing the respondent’s wine grapes as the source of the appellants’ wine

can therefore safely be excluded. [17] The only enquiry worthy of pursuit therefore relates to the likelihood of notional purchasers of Zonquasdrift wine believing that they originate from the same farm as the respondent’s grapes. The first question raised by this enquiry is whether the notional wine purchaser would be aware of the respondent’s grapes that are sold under its trade name. Since the respondent does not market its wares in retail outlets or advertise them in the public domain, I cannot see why this awareness should be assumed. My second problem with the appellants’ argument is this: in order for the notional purchaser to be confused in thinking that Zonquasdrift wine comes from the same farm as Zonquasdrif grapes, he or she must infer that (a) there is a farm Zonquasdrift; (b) that farm is owned by the same entity as the trade mark; and (c) that the wine is made of grapes grown on that farm. As I see it, however, there is no reason to assume that the notional purchaser

will make all these inferences just by looking at the appellants’ trade mark on their wine label. [18] As to (a): it is well known that many trade marks in the wine industry are not associated with farms at all. That is so even where the trade mark bears resemblance to the name of an existing or non-existing farm. As to (b): even if the notional wine purchaser knew of the farm Zonquasdrift, there is no reason to think that he or she would have known or infer that the farm and the trade name belong to the same proprietor. As to (c): even if the notional purchaser knows or infers that there is a farm Zonquasdrift which belongs to the same entity as the trade mark, he or she would have no reason to infer that Zonquasdrift wine is made of grapes grown on that farm. An interesting exchange on the papers illustrates this

point. In the answering affidavit Walters contended that, since the appellants’ Zonquasdrift wine is made by Riebeek

Cellars that sources its grapes from 32 farms, it cannot even be said that the wine sold under the appellants’ trade mark is made from grapes grown on its farm. In reply the appellants then said – and this was confirmed by Mr Izak Bester who is the wine-maker at Riebeek Cellars - that Zonquasdrift wine is indeed made from chenin blanc grapes grown on the appellants' farm. The moral of the story is this: although there is no reason to doubt the correctness of Bester’s version, he was clearly one of a very few who could say where the grapes came from. For all the potential purchaser knows, the grapes could come from one or more of 32 farms. [19] Even if the appellants were therefore to make their own wine on the farm Zonquasdrif, they would not be obliged to source the grapes for wine-making under their trade mark from that farm only. In consequence the customer would have no basis to think that they did. According to well-established principle, a trade mark serves as a badge of origin, in the sense that it identifies and guarantees the trade origin of the goods to which it applies (see eg Verimark (Pty) Ltd v BMW AG 2007 (6) SA 263 (SCA) paras 4-5). But in the wine industry a trade mark in respect of wine serves to guarantee the origin of the wine only. The badge of origin in respect of the grapes from which that wine is made is provided by the wine of origin

scheme (WO scheme). [20] The WO scheme is governed by regulations promulgated (under GNR 1434 of 29 June 1990) in terms of s 14 of the Liquor Products Act 60 of 1989. It emanates from the recognition in the wine industry that the ‘terroir’ or area where the grapes are grown has a significant bearing on the wine made from these grapes. Under the WO scheme geographical areas in which wine are produced are divided into nine units. These units are further divided into increasingly smaller areas called ‘regions’,

‘districts’ and ‘wards’. The farms of the appellants and the respondent, for instance, fall into the Swartland region and the Riebeek Kasteel district. As to the origin of the grapes from which the wine is made, the potential purchaser can therefore rely on the designation of origin on the label. Since the appellants’ wine is designated as a wine of origin Swartland, the purchaser can assume the grapes are grown in that area. But with regard to the origin of the grapes he or she can assume nothing more. For these reasons I find no substance in the appellants’ contention that they have established a likelihood of confusion with regard to the origin of their wine and the respondent’s grapes. [21] In further support of their case, the appellants sought to rely on the fact that at some stage Riebeek Cellars sold its ‘Limited Release Chenin Blanc’ with an indication on the label that the grapes came from the single vineyard on the farm Zonquasdrift. But as I see it, this reliance is misconceived. The wine was sold under the trade mark of Riebeek Cellars. The reference to Zonquasdrift related to appellants’ farm – not their trade mark – as the origin of the grapes which, as we know, is not covered by their registered mark. [22] The appellants’ further argument rested on two alleged incidents of actual confusion. The first arose when Eskom threatened to cut off the electricity supply

to the appellants’ farm because the respondent had failed to pay its account. The second occurred when the appellants’ farm manager was asked by a third party whether the appellants had opened a new business under the name Zonquasdrift Vineyards CC. But I find the argument based on these instances equally unsustainable. The confusion demonstrated by these incidents, as I see it, probably arose from the similarity between the respondent’s name and that of the appellants’ farm. It therefore had nothing to do with confusion created by two similar marks. [23] This brings me to the appellants’ application in terms of s 20(2)(b) of the Close Corporations Act, to compel the change of the respondent’s close corporation name on the basis that it is calculated to cause damage to the appellants. As it turned out, this part of the case does not require any lengthy discourse. The reason for this emanates from the narrow foundation upon which the appellants eventually relied as the basis for this relief. All that the appellants contended for in the end was that, because the use of the respondent’s name results in an infringement of their trade mark, it is calculated to cause damage to them. Since I hold the view that the infringement of the appellants’ mark had not been established, this application must follow the same fate. [24] What remains is the appeal against the costs order relating

to the respondent’s counter-application. It will be remembered that the court a quo dismissed the counter-application with no order as to costs. The appellants’ objection against this order starts out from the premise that, as a rule, the dismissal of an application results in a costs order against the unsuccessful applicant. That is undoubtedly so. But it is not a rule of the Medes and Persians. In this case the court a quo found it unnecessary to deal with the counter-application because it rightly regarded it as no more than a second tier defence against the main application which was dismissed on the basis of the respondent’s first tier defence. In the exercise of its discretion the court a quo then decided that, in these circumstances, the most appropriate order would be no order as to costs. According to a well-established principle, this court will not interfere with the decision of a lower court as to costs unless it is quite clear that some important factor escaped the attention of the lower court, or unless the discretion exercised has not been a judicial discretion (see eg Molteno Bros v South African Railways 1936 AD 408 at 417). In this light, I can see no basis upon which we can interfere with the exercise of the court a quo’s discretion with regard to costs. [25] As to the costs on appeal, the respondent was at all times represented by two counsel. Shortly before the hearing of the appeal, its senior counsel however became indisposed. That is the reason for the costs order I propose to make. In the result: The appeal is dismissed with costs, including the costs of two counsel whenever employed.

F D J BRAND

JUDGE OF APPEAL APPEARANCES: For Appellants: D HARMS Instructed by: ADAMS & ADAMS

CAPE

TOWN Correspondents:

HONEY ATTORNEYS

BLOEMFONTEIN For Respondent: M IOANNOU (HEADS OF ARGUMENT PREPARED BY A R SHOLTO-DOUGLAS SC AND M IOANNOU) Instructed by: MacROBERT INC

CAPE

TOWN Correspondents:

CLAUDE REID

BLOEMFONTEIN

THE SUPREME COURT

OF APPEAL OF SOUTH AFRICA

JUDGMENT

REPORTABLE

Case No: 965/12

In the matter between:

ALEXANDER

ANATOLE THEODOR METTENHEIMER ..................................FIRST

APPELLANT

ZONQUASDRIFT ESTATES (PTY) LIMITED ...............................................SECOND

APPELLANT

and

ZONQUASDRIF

VINEYARDS CC ..................................................................FIRST

RESPONDENT

REGISTRAR OF CLOSE CORPORATIONS ............................................SECOND

RESPONDENT

REGISTRAR OF TRADE

MARKS .................................................................THIRD

RESPONDENT

Neutral citation: Mettenheimer v Zonquasdrif Vineyards CC (965/12) [2013] ZASCA 152 (19 November 2013)

Coram: Brand, Theron, Pillay, Petse JJA et Meyer AJA

Heard: 4 November 2013

Delivered: 19 November 2013

Summary: (a) Trade Marks Act 194 of 1993 – appellants’ trade mark registered in respect of wine – used by the respondent in the course of trade with reference to wine grapes – whether goods so similar as to give rise to the likelihood of deception or confusion as contemplated in s 34(1)(b) of the Act.

(b) Close Corporations Act 69 of 1984 – whether respondent’s name calculated to cause damage to the appellants as contemplated in s 20(2)(b) of the Act.

ORDER

On appeal from: Western Cape High Court, Cape Town (Davis J sitting as court of first instance):

The appeal is dismissed with costs, including the costs of two counsel whenever employed.

BRAND JA (THERON, PILLAY, PETSE JJA et MEYER AJA concurring):

[1] This appeal turns on the application of s 34(1)(b) of the Trade Marks Act 194 of 1993 and s 20(2)(b) of the Close Corporations Act 69 of 1984 prior to its amendment that came into operation in 2011. The first appellant, Mr Alexander Mettenheimer, is the registered proprietor of the trade mark Zonquasdrift in class 33 in respect of alcoholic beverages, except beer. The trade mark therefore covers wine but not wine grapes which, incidentally, falls in class 31. Mettenheimer and his wife are also the shareholders in the second appellant, a private company which is the owner of a farm called Zonquasdrift between Malmesbury and Riebeek Kasteel in the Western Cape. Until recently the second appellant was known as Almett Properties (Pty) Ltd, but on 3 May 2012 – that is about two years after the commencement of the present litigation – it formally changed its name to Zonquasdrift Estates (Pty) Ltd.

[2] The first respondent is a close corporation with the registered name Zonquasdrif Vineyards CC. It conducts its farming business on a farm situated about one kilometre from the second appellant’s farm where it grows wine grapes which it sells under its registered name. The second and third respondents are the Registrar of Close Corporations and the Registrar of Trade Marks, respectively, who were joined in the proceedings for the sake of formality and who never became active participants at any stage. Hence I propose to refer to the two appellants as ‘the appellants’ and to the first respondent as ‘the respondent’.

[3] Proceedings started when the appellants brought an application against the respondent in the Western Cape High Court, Cape Town. The relief they sought was twofold. First, that the respondent be interdicted in terms of s 34(1)(b) of the Trade Marks Act from infringing its trade mark Zonquasdrift by, inter alia, selling wine grapes under that name. Secondly, for an order declaring that the respondent’s registered name is calculated to cause damage to the appellants and ordering the respondent, for that reason, to change its name in terms of s 20(2)(b) of the Close Corporations Act. The respondent not only opposed the relief sought on its merits, but also brought a counter-application for the removal of the appellants’ trade mark from the register on the basis that it exclusively serves to ‘designate the geographical area of the goods’ as envisaged by s 10(2)(b) of the Trade Marks Act.

[5] The issues that arose on appeal will be better understood against the background facts that follow. I find it convenient to start with the geographical layout of the area involved. Zonquasdrift is the name of a natural crossing over the Berg River or a ‘drift’ in Dutch, which eventually became a ‘drif’ in Afrikaans. Historians tell us that the existence of the crossing has been well documented under that name since at least 1660 when the Dutch colonists first used it to cross the natural hindrance to their passage into the interior, created by the Berg River. In more modern times Zonquasdrift has also become known as the end of the first stage of the four day Berg River canoe marathon. The farm Zonquasdrift was originally granted in the early 1700s. Rather self-evidently, the crossing is on that original grant.

The farm of the appellants is that subdivision of the original grant which retained the original name. On the one side of the

Berg River the crossing is now on the appellants’ farm. On the opposite side of the river the crossing is on the farm Klein Zonquasdrift. The farm on which the respondent conducts its business is on the same side of the river as Klein Zonquasdrift. It is about 300 metres from the crossing and, as I have said, about one kilometre from the appellants’ farm. It is a subdivision of the farm Bosplaas. Until it was given the name Zonquasdrif Vineyards in 2008, it was never known by that name.

[6] The appellants are not involved in wine-making. They do not have a cellar on Zonquasdrift farm nor do they employ any wine-makers. It is a wine grape producing farm that delivers its grapes in bulk to a co-operative cellar, Riebeek Cellars. The vinification process is conducted exclusively by Riebeek Cellars, which sources its wine grapes from 32 different farms in the Swartland area, being the members of the co-operative. The wine is then sold under the trade mark Riebeek Cellars. As a result of this communal system, it is generally not possible to say, with reference to a particular wine, where the grapes originally came from. The grapes are pooled by Riebeek Cellars and the wine sold under its label with the designation that it is a wine of origin Swartland under the wine of origin scheme, to which I shall presently return.

[7] An exception to the general exposition conveyed thus far, pertains to wine made by Riebeek Cellars from grapes of the chenin blanc cultivar that originates from two blocks on the appellants’ farm – about six hectares in extent – which had been registered under the wine of origin scheme as a single vineyard. Pursuant to an arrangement between the appellants and Riebeek Cellars, chenin blanc wine from this single vineyard is kept separate and bottled under the special Zonquasdrift label. This wine is then exported by the appellants for sale in Germany. It also bears the designation wine of origin Swartland. Moreover Riebeek Cellars has in 2009 bottled a chenin blanc wine from the appellants’ single vineyard as ‘Limited Release Chenin Blanc’ under the Riebeek Cellars trade mark, but with the indication on

the label that it comes ‘from a designated block of vines in a 47-year old single vineyard on the farm Zonquasdrift in the Swartland’. In short, the appellants’ business is therefore that of a wine grape producer and on a more limited scale, that of a wine exporter under its registered trade mark. What should perhaps be pointed out for the sake of completeness, is that the appellants never registered any trade mark for their wine grapes in class 31.

[8] The respondent was incorporated on 4 August 2008. Its majority member and the deponent to its answering affidavit is Mr Petrus Walters. In the same way as the appellants, the respondent is also a wine grape producer. But, unlike the appellants, it has never sold wines under any name. Neither has it ever made any wine. It sells its wine grapes in bulk under its registered name, Zonquasdrif Vineyards, to a corporate wine cellar in the Worcester area. By contrast, the farm Klein Zonquasdrift, which is directly across the Berg River from the appellants’ farm, produces table grapes. Table grapes are not used to produce wine. Conversely, wine grapes are not suitable for consumption as a fruit in the way that table grapes are. In consequence, wine grapes are not sold to the general public, but only to wine cellars or wine co-operatives. As to why the respondent was given the corporate and trading name Zonquasdrif Vineyards CC, Walters explained that he found the original name of the farm on which the respondent conducts business, Bosplaas, rather uninspiring. In consequence he decided to link its name to the historical nearby crossing, which is well-known to the public, so as to indicate the geographical location of the respondent’s farming business.

[9] In delineating the issues it is perhaps useful to identify that which falls outside the ambit of this appeal. The appellants did not rely on passing off or any other form of unlawful competition to justify any of the relief sought in the court a quo. The case relied solely on an infringement of their trade mark rights. Potential confusion resulting from similar names of farms situated in close proximity to one another, for instance, is therefore of no consequence. Equally irrelevant is the respondent’s contention that the appellants had failed to establish a reputation either to its grape production or its wine. The fact that the appellants thus far have only sold their wine in Germany, is not material. What we have to look at is the ‘notional use’ that the appellants can make of their trade mark, which would obviously include the marketing of wine under that label in South Africa (see eg Plascon-Evans Paints Ltd v Van Riebeeck Paints (Pty) Ltd [1984] ZASCA 51; 1984 (3) SA 623 (A) at 641E-I).

[10] The interdict application, as I have said by way of introduction, rested squarely on s 34(1)(b) of the Trade Marks Act. In relevant part this section provides:

‘(1) The rights acquired by registration of a trade mark shall be infringed by-

(a) . . .

(b) the unauthorized use of a mark which is identical or similar to the trade mark registered, in the course of trade in relation to goods or services which are so similar to the goods or services in respect of which the trade mark is registered, that in such use there exists the likelihood of deception or confusion.’

[11] The respondent clearly used its impugned mark in the course of trade and it is not suggested that it had been authorised by the appellants to do so. That limits the enquiry to the issue of confusing similarity. Unlike s 34(1)(a), the provisions of s 34(1)(b) do not require that the offending mark be used in relation to goods in the class for which the trade mark had been registered.

It contemplates two elements, namely, (a) a mark identical or similar to the trade mark used in relation (b) to goods which are so similar to those for which it had been registered, that it gives rise to a likelihood of deception or confusion. As to the relationship between these two elements, I agree with the sentiments expressed by Thring J when he said in New Media Publishing (Pty) Ltd v Eating Out Web Services CC [2005] ZAWCHC 20; 2005 (5) SA 388 (C) at 394C-F:

‘There is, it seems to me, an interdependence between the two legs of the inquiry: the less the similarity between the respective goods or services of the parties, the greater will be the degree of resemblance required between their respective marks before it can be said that there is a likelihood of deception or confusion in the use of the allegedly offending mark, and vice versa. Of course, if the respective goods or services of the parties are so dissimilar to each other that there is no likelihood of deception or confusion, the use by the respondent even of a mark which is identical to the applicant's registered mark will not constitute an infringement; also, if the two marks are sufficiently dissimilar to each other no amount of similarity between the respective goods or services of the parties will suffice to bring about an infringement. . . .’

(See also Canon Kabushiki Kaisha v Metro-Goldwyn-Mayer Inc [1999] RPC 117 (ECJ) para 17.)

[12] In comparing the two marks I find them virtually identical. Although the respondent’s mark contains the additional

reference to ‘Vineyards’ I regard the dominant or striking feature of the mark, in the sense that it will impact on the mind and recollection of the notional ordinary purchaser, in ‘Zonquasdrif’ (see eg Plascon-Evans Paints Ltd v Van Riebeek Paints (Pty) Ltd [1984] ZASCA 51; 1984 (3) SA 623 (A) at 641D-E). The further difference between the two, which lies in the absence of the final ‘t’ in the respondent’s mark – changing it from Dutch into Afrikaans – is in my view insignificant. In the unlikely event that it is noticed by the notional purchaser, it could be ascribed to a misspelling of the appellants’ mark. What it therefore boils down to in the end is the similarity of the goods. Can it be said that, having regard to the sameness of the two marks, the similarity between the goods in respect of which the appellants’ mark is registered (wine) and the goods in

which respondent trades (wine grapes) is such that confusion or deception is the probable result? The deception and confusion we are talking about, of course, must relate to the origin of the respective goods. So, is it likely that the notional purchaser may be confused to think that these goods have the same origin?

[13] Considerations that could assist in determining this likelihood of confusion have been proposed in decided cases. Prominent

amongst these are the judgments of the Chancery Division in British Sugar Plc v James Robertson & Sons Ltd [1996] RPC 281 (Ch D) at 296-297 and of this court, albeit in a context different from s 34(1)(b) in Danco Clothing (Pty) Ltd v Nu-Care Marketing Sales and Promotions (Pty) Ltd [1991] ZASCA 121; 1991 (4) SA 850 (A) at 860-861. Included amongst the considerations proposed in these cases are: (a) the uses of the respective goods; (b) the users of the respective goods; (c) the physical nature of the goods; and (d) the respective trade channels through which the goods reach the market. It is clear that the list of proposed considerations was never intended to be exhaustive. Yet they serve to stimulate and guide the required process of reasoning in the enquiry as to the likelihood of confusion.

[14] On application of these considerations to wine grapes and wine, first impressions are that the likelihood of confusion is slight indeed. The nature of the two products is entirely different. The one is a fruit – albeit inedible – and the other is an alcoholic beverage. As are their uses, their users and the trade channels through which they are marketed. Since wine grapes are not suitable for consumption as a fruit, they are not sold to the public and they are therefore not to be found in any retail outlets. Wine, on the other hand, is marketed, advertised and sold directly to the public in supermarkets, liquor stores and other retail outlets. The prospects of Zonquasdrift wine and Zonquasdrif grapes ever being marketed or sold in close proximity can therefore safely be excluded as non-existent.

[15] The appellants’ argument as to why, despite these distinct differences between the two products, they have succeeded in showing a likelihood of confusion, went along the following lines. Wine grapes constitute the raw material from which wine is made. Hence the two products are associated with one another. It is known in the industry that while some farms sell their wine grapes to wineries, others make their own wine from grapes produced on the farm or purchased from other farms. There is therefore no clear delineation between farms producing and selling grapes and those selling both wine and grapes. Although the appellants do not make their own wine and only sell their Zonquasdrift wine in Germany, they are entitled to sell their wine in South Africa as well. What must also be postulated as part of the appellants’ notional use

is that they may integrate their grape growing and wine-making activities on the farm Zonquasdrift. In these circumstances, so the appellants’ argument went, the notional member of the public may very well be confused in thinking that the appellants’ wine and the respondent’s grapes originate from the same farm.

[16] In evaluating the appellants’ argument, it seems to me that it draws no distinction between the notional purchasers of grapes, on the one hand, and the notional purchasers of wine, on the other. But, in the light of the distinct differences between the two products that I have alluded to, I do not believe logical reasoning allows this conflation. For the very reason that wine grapes are suitable for wine-making only, they are exclusively sold to wine makers and wine co-operatives for that purpose. The undisputed evidence by the respondent is that wine cellars and wineries purchase grapes based on a number of factors, including the cultivar and the terroir, the vineyards and the like but not with reference to the trade names of wine or the names of farms. Furthermore, wineries and co-operative cellars make use of specialists in the wine industry to buy their wine grapes from approved suppliers who comply with their own established quality control standards. The chances of these buyers confusing the respondent’s wine grapes as the source of the appellants’ wine can therefore safely be excluded.

[17] The only enquiry worthy of pursuit therefore relates to the likelihood of notional purchasers of Zonquasdrift wine believing that they originate from the same farm as the respondent’s grapes. The first question raised by this enquiry is whether the notional wine purchaser would be aware of the respondent’s grapes that are sold under its trade name. Since the respondent does not market its wares in retail outlets or advertise them in the public domain, I cannot see why this awareness should be assumed. My second problem with the appellants’ argument is this: in order for the notional purchaser to be confused in thinking that Zonquasdrift wine comes from the same farm as Zonquasdrif grapes, he or she must infer that (a) there is a farm Zonquasdrift; (b) that farm is owned by the same entity as the trade mark; and (c) that the wine is made of grapes grown on that farm. As I see it, however, there is no reason to assume that the notional purchaser

will make all these inferences just by looking at the appellants’ trade mark on their wine label.

[18] As to (a): it is well known that many trade marks in the wine industry are not associated with farms at all. That is so even where the trade mark bears resemblance to the name of an existing or non-existing farm. As to (b): even if the notional wine purchaser knew of the farm Zonquasdrift, there is no reason to think that he or she would have known or infer that the farm and the trade name belong to the same proprietor. As to (c): even if the notional purchaser knows or infers that there is a farm Zonquasdrift which belongs to the same entity as the trade mark, he or she would have no reason to infer that Zonquasdrift wine is made of grapes grown on that farm. An interesting exchange on the papers illustrates this

point. In the answering affidavit Walters contended that, since the appellants’ Zonquasdrift wine is made by Riebeek

Cellars that sources its grapes from 32 farms, it cannot even be said that the wine sold under the appellants’ trade mark is made from grapes grown on its farm. In reply the appellants then said – and this was confirmed by Mr Izak Bester who is the wine-maker at Riebeek Cellars - that Zonquasdrift wine is indeed made from chenin blanc grapes grown on the appellants' farm. The moral of the story is this: although there is no reason to doubt the correctness of Bester’s version, he was clearly one of a very few who could say where the grapes came from. For all the potential purchaser knows, the grapes could come from one or more of 32 farms.

[19] Even if the appellants were therefore to make their own wine on the farm Zonquasdrif, they would not be obliged to source the grapes for wine-making under their trade mark from that farm only. In consequence the customer would have no basis to think that they did. According to well-established principle, a trade mark serves as a badge of origin, in the sense that it identifies and guarantees the trade origin of the goods to which it applies (see eg Verimark (Pty) Ltd v BMW AG 2007 (6) SA 263 (SCA) paras 4-5). But in the wine industry a trade mark in respect of wine serves to guarantee the origin of the wine only. The badge of origin in respect of the grapes from which that wine is made is provided by the wine of origin scheme (WO scheme).

[20] The WO scheme is governed by regulations promulgated (under GNR 1434 of 29 June 1990) in terms of s 14 of the Liquor Products Act 60 of 1989. It emanates from the recognition in the wine industry that the ‘terroir’ or area where the grapes are grown has a significant bearing on the wine made from these grapes. Under the WO scheme geographical areas in which wine are produced are divided into nine units. These units are further divided into increasingly smaller areas called ‘regions’,

‘districts’ and ‘wards’. The farms of the appellants and the respondent, for instance, fall into the Swartland region and the Riebeek Kasteel district. As to the origin of the grapes from which the wine is made, the potential purchaser can therefore rely on the designation of origin on the label. Since the appellants’ wine is designated as a wine of origin Swartland, the purchaser can assume the grapes are grown in that area. But with regard to the origin of the grapes he or she can assume nothing more. For these reasons I find no substance in the appellants’ contention that they have established a likelihood of confusion with regard to the origin of their wine and the respondent’s grapes.

[21] In further support of their case, the appellants sought to rely on the fact that at some stage Riebeek Cellars sold its ‘Limited Release Chenin Blanc’ with an indication on the label that the grapes came from the single vineyard on the farm Zonquasdrift. But as I see it, this reliance is misconceived. The wine was sold under the trade mark of Riebeek Cellars. The reference to Zonquasdrift related to appellants’ farm – not their trade mark – as the origin of the grapes which, as we know, is not covered by their registered mark.

[22] The appellants’ further argument rested on two alleged incidents of actual confusion. The first arose when Eskom threatened to cut off the electricity supply to the appellants’ farm because the respondent had failed to pay its account. The second occurred when the appellants’ farm manager was asked by a third party whether the appellants had opened a new business under the name Zonquasdrift Vineyards CC. But I find the argument based on these instances equally unsustainable. The confusion demonstrated by these incidents, as I see it, probably arose from the similarity between the respondent’s name and that of the appellants’ farm. It therefore had nothing to do with confusion created by two similar marks.

[23] This brings me to the appellants’ application in terms of s 20(2)(b) of the Close Corporations Act, to compel the change of the respondent’s close corporation name on the basis that it is calculated to cause damage to the appellants. As it turned out, this part of the case does not require any lengthy discourse. The reason for this emanates from the narrow foundation upon which the appellants eventually relied as the basis for this relief. All that the appellants contended for in the end was that, because the use of the respondent’s name results in an infringement of their trade mark, it is calculated to cause damage to them. Since I hold the view that the infringement of the appellants’ mark had not been established, this application must follow the same fate.

[24] What remains is the appeal against the costs order relating to the respondent’s counter-application. It will be remembered that the court a quo dismissed the counter-application with no order as to costs. The appellants’ objection against this order starts out from the premise that, as a rule, the dismissal of an application results in a costs order against the unsuccessful applicant. That is undoubtedly so. But it is not a rule of the Medes and Persians. In this case the court a quo found it unnecessary to deal with the counter-application because it rightly regarded it as no more than a second tier defence against the main application which was dismissed on the basis of the respondent’s first tier defence. In the exercise of its discretion the court a quo then decided that, in these circumstances, the most appropriate order would be no order as to costs. According to a well-established principle, this court will not interfere with the decision of a lower court as to costs unless it is quite clear that some important factor escaped the attention of the lower court, or unless the discretion exercised has not been a judicial discretion (see eg Molteno Bros v South African Railways 1936 AD 408 at 417). In this light, I can see no basis upon which we can interfere with the exercise of the court a quo’s discretion with regard to costs.

[25] As to the costs on appeal, the respondent was at all times represented by two counsel. Shortly before the hearing of the appeal, its senior counsel however became indisposed. That is the reason for the costs order I propose to make. In the result:

F D J BRAND

JUDGE OF APPEAL

APPEARANCES:

For Appellants: D HARMS

Instructed by: ADAMS & ADAMS

CAPE

TOWN

Correspondents:

HONEY ATTORNEYS

BLOEMFONTEIN

For Respondent: M IOANNOU

(HEADS OF ARGUMENT PREPARED BY A R SHOLTO-DOUGLAS SC AND M IOANNOU)

Instructed by: MacROBERT INC

Correspondents:

CLAUDE REID