South Africa: Supreme Court of Appeal

REPUBLIC OF SOUTH AFRICA

IN THE SUPREME COURT OF APPEAL

OF SOUTH AFRICA

REPORTABLE

Case number: 344/99

In the matter between:


RENTREAG MARKETING (PTY) LTD 1^st Appellant
PATRICK LORENZ GAERTNER 2^nd Appellant
RORY CHARLES KLEMP 3^rd Appellant


and


THE COMMISSIONER OF CUSTOMS AND EXCISE Respondent



CORAM: SCHUTZ JA, MELUNSKY and NUGENT AJJA

HEARD: 20 FEBRUARY 2001

DELIVERED: 22 MARCH 2001

___________________________________________________________________

JUDGMENT

__________________________________________________________________

MELUNSKY AJA:
[1] What has to be decided in this appeal is whether certain consignments of cheese imported into South Africa from Australia were properly classified as Gouda by the respondent, the Commissioner for Customs and Excise (“the Commissioner”) in terms of the Customs and Excise Act 91 of 1964 (“the Act”). The cheese was imported by E M Gaertner Trading CC (“the corporation”) which was subsequently converted into a company, Rentreag Marketing (Pty) Ltd (“Rentreag”). Although Rentreag was originally the first appellant in the appeal it has since been wound up and the liquidator abides by the decision of this Court. The remaining appellants, Mr Patrick Gaertner and Mr Rory Klemp were members of the corporation at the material time.
[2] During the latter part of 1996 the corporation commenced importing cheese from Australia and Canada in containers. The Australian cheese was produced by Lactos (Pty) Ltd (“Lactos”) and the Canadian cheese was manufactured by Saputo Limited, Quebec. According to Gaertner the corporation regarded the cheese as Edam and, for the purposes of importing it into South Africa, described it as “other” under tariff sub-heading 0406.90.90 which forms part of Schedule 1 to the Act. For the purposes of this appeal the relevant sub-heading is 0406.90 which is divided into four categories attracting different rates of duty. These are:

There have since been substantial amendments to the duties payable but these are of no consequence for present purposes.
[3] The Commissioner determined that all of the cheese was Gouda and that it attracted a duty of 660c per kg under tariff sub-heading 0406.90.35, which is substantially higher than the rate of duty payable in respect of “other”. If the Commissioner’s determinations are correct, it is accepted that Gaertner and Klemp are personally liable in terms of s103 of the Act for the payment of additional duty under sub-heading 0406.90.35.
[4] The Commissioner’s determinations were taken on appeal on notice of motion to the Transvaal Provincial Division in terms of s 47(9)(e) of the Act. Botha J was of the view that both the Australian and the Canadian cheese fell under tariff heading 0406.90.90 and he granted an order to this effect. The Commissioner’s appeal to the full court was successful in so far as it related to the cheese imported from Australia. Van Dijkhorst J, with whom Stafford DJP and Van Dyk J concurred, set aside Botha J’s order and replaced it with the following:

“1. An order is granted in terms of Section 47(9) of the Customs and Excise Act 91 of 1964 that the written tariff determinations of the commissioner of customs and excise in respect of the consignments of cheese imported by E M Gaertner Trading CC which are referred to in letters by the controller of customs and excise Cape Town dated 17 June 1997 and by the commissioner for customs and excise dated 13 August 1997 in so far as they relate to Canadian cheese produced by Saputo Ltd, Quebec are corrected by substituting therefor determinations to the effect that the said consignments fall under tariff heading 0406.90.90 in chapter IV of schedule I to the said act.

Each party will pay its own costs.”

The effect of the order was that the Commissioner’s determination of the Australian cheese as Gouda remained effectual. This is an appeal against the order of the full court with the special leave of this Court.
[5] Before dealing with the essential question for decision, two preliminary matters may be disposed of. The first is that both parties accepted that all of the cheese imported from Australia has identical characteristics and that the samples analysed and tested are representative of the whole. It may be noted, too, that although some of the cheese was orange-coloured and the remainder had a natural colour, it was not argued that the difference in colour had any material bearing on the classification, despite the fact that one of the expert witnesses did notice a difference in taste between the two. Secondly, and while the appellants initially contended that the cheese belonged to the variety known as Edam, they conceded in the court of first instance that they were not able to identify the cheese positively but that it nevertheless fell under tariff sub-heading 0406.90.90. In this Court the Commissioner’s counsel accepted that the appellants’ failure to categorise the cheese did not preclude us from setting aside the Commissioner’s determination if it was established that the Australian cheese was not Gouda. It was submitted, however, that in this event the matter should be referred back to the Commissioner to enable him to reconsider his determination.
[6] Despite the many and lengthy affidavits before us, the outcome of the appeal depends largely on a proper consideration of the evidence of two expert witnesses - Prof Walstra and Ms Wessels. It was not disputed that the word “Gouda” is a special or technical term and that it should be interpreted in the light of the evidence of persons skilled or knowledgeable in the field of cheeses (see Letraset Ltd v Helios Ltd 1972 (3) SA 245 (A) at 250 C-D and International Business Machines SA (Pty) Ltd v Commissioner for Customs and Excise 1985 (4) SA 852 (A) at 874 B). It may be added that an appeal in terms of s 47(9)(e) of the Act is an appeal in the wide sense, that is a complete rehearing on the merits with or without evidence (see Tikly and Others v Johannes NO and Others 1963 (2) SA 588 (T) at 590 G-H and National Union of Textile Workers v Textile Workers Industrial Union (SA) and Others 1988 (1) SA 925 (A) at 937 D-G). This is so because the Act does not envisage a formal hearing by the Commissioner although, before making a determination, he may be bound to exercise fair administrative procedures in terms of s 33 of the Constitution. Consequently it is permissible for the parties to introduce additional evidence that is relevant to the issues.
[7] Walstra, whose affidavits were filed on the appellants’ behalf, is particularly well-qualified in his field. He is an emeritus professor of the Agricultural University of Wageningin in the Netherlands, where he has taught for almost forty years in the field of the chemistry and physics of dairy products, including cheese. He is the author of about 160 publications in scientific journals and is a contributor to books dealing with dairy products. Several of his publications relate to various aspects of the manufacture and properties of cheese. It is clear that he is one of the leading experts on Dutch-type cheeses, including, of course, Gouda. Wessels is the quality manager of the Animal Nutrition and Animal Products Institute (“the ANPI”), a division of the Agricultural Research Council situated in Pretoria. She has had extensive experience in the field of dairy science and, apart from other positions held by her, she is the President of the South African Society of Dairy Technology (Transvaal Division). She has considerable knowledge of dairy and cheese products and is qualified to carry out and supervise tests on such products. Counsel for the appellants referred to her, perhaps somewhat dismissively, as a mere technician but this criticism is not justified. Wessels clearly does not have the experience of Walstra nor his depth of knowledge of cheeses, especially those from the Netherlands, but she is sufficiently qualified to express an opinion on whether or not cheese falls within the category of Gouda. It may be noted in this respect that counsel were agreed that Gouda (or “sweetmilk” as it is sometimes known) is one of the cheeses most frequently bought and consumed in South Africa - hence the higher duty that prevailed at the time.

[8] A second criticism of Wessels’s opinions related to her report of 10 June 1997 in which she said that the cheese in question was not Edam because, according to the Marketing Act 53 of 1986, the percentage of fat in dry matter (“FIDM”) in respect of Edam was specified as being between 40 and 45, whereas the percentage of FIDM for Gouda was required to be between 48 and 60. (Wessels’s reference to the Marketing Act 53 of 1986 was erroneous. She obviously intended to refer to the regulations relating to dairy products made under the Marketing Act 59 of 1968 and deemed to be in force under the Agricultural Product Standards Act 119 of 1990.) The cheese which she tested had a percentage of FIDM between 48 and 60 and she concluded that it was therefore Gouda. Counsel for the appellants pointed out the there are at least 25 other cheese varieties with a FIDM percentage between 48 and 60 and to classify the cheese in question as Gouda solely on that ground was not a proper scientific evaluation. However it would seem that Wessels’s report was furnished in response to a query directed to the ANPI by the Commissioner who wanted to know whether the cheese was Edam, as the corporation then claimed, or whether it was Gouda, as the Commissioner suspected it might be. As the high percentage of FIDM content clearly disqualified the cheese from being classified as Edam but as the percentage fell within the Gouda range, Wessels concluded that it was indeed Gouda. It is generally accepted world-wide that, in order to be classified as Gouda, the fat content as expressed as a percentage of the dry matter must be between 48 and 60, whereas the range for Edam is between 40 and 45. The appellants agree that the percentage of FIDM in the imported Australian cheese fell within the Gouda range and for this reason alone the cheese could not be classified as Edam.
[9] Counsel for the appellant relied on the affidavit of Mr Michael White, the export and industrial sales manager of Lactos, in order to establish that Lactos had intended to produce Edam and not Gouda for export to the corporation. White stated that although Lactos also produced Gouda, it used a different process to do so. One of the questions raised during argument was whether it is permissible for the Commissioner - or the Court on appeal - to have regard to the manufacturing process for the purpose of determining under which tariff heading imported goods should be classified. Very little argument was addressed to us on the point and, as far as I am aware, this is not a matter which has previously been pronounced upon by this Court. Without objection from counsel for the Commissioner, counsel for the appellants addressed us on the process that should be employed to produce Gouda cheese and the differences between the Edam and Gouda processes in the Lactos factory. Moreover both Walstra and Wessels covered this ground in arriving at their respective conclusions. For the purposes of this appeal, therefore, it will be assumed that we are not precluded from considering the process of manufacture of Gouda or the process that was in fact used by Lactos.
[10] There are three significant differences between the views of Walstra and those of Wessels. The first concerns the absence of “eyes” in the imported cheese, the second relates to the temperature and period for the maturation of Gouda and the third to the consistency and flavour of the cheese in question. Walstra also mentioned that the rind of the imported cheese was thin and soft, unlike the rind of a traditional Gouda but he accepted that Gouda with a thin, soft rind is also produced currently. Walstra and Wessels were in agreement that in the manufacture of Gouda it is usual to use cows’ milk pasteurised at 72.5 degrees Celsius for approximately 15 seconds. This, it may be noted, was the type of milk used by Lactos in the manufacture of the cheese imported by the corporation.
[11] According to Walstra it is the type of mesophyllic starter culture (lactic acid bacteria) normally used for the manufacture of Gouda that produces carbon dioxide which, in turn, causes round openings or eyes in the body of the cheese. Frank Kosikowski, the author of an authoritative text book dealing with the manufacture and characteristics of Gouda and Edam, says that the presence of shiny eyes in Gouda is considered to be normal. Although Wessels associated herself “fully” with Kosikowski’s views, she denied that the mesophyllic starter culture which produces carbon dioxide, and thus eyes, is normally used to produce Gouda, according to current manufacturing processes. She added that it is not uncommon to find a Gouda cheese with no holes at all. The International Dairy Federation (“the IDF”), an authoritative international body, has from time to time provided a catalogue of cheeses “to promote the fair use of cheese designations in international trade”. The most recent (1981) catalogue reflects that eyes are present in almost all Gouda cheese, including South African Gouda. Australian Gouda, according to the catalogue, has “medium sized round openings”. Significantly enough, however, there are no openings or eyes in the interior of French and Japanese Gouda. Moreover Wessels pointed out that since 1981 the methods and techniques of cheese manufacture have changed considerably and that the characteristics specified in the IDF catalogue do not necessarily continue to apply in all respects.
[12] Although the existence of eyes may be normal or usual in Gouda, the evidence falls short of establishing that their presence is essential for the cheese to be classified as such. Quite apart from the fact that there are no eyes in French or Japanese Gouda, Walstra himself mentions that the mesophyllic starter culture normally used in the manufacture of Gouda produces carbon dioxide and thus holes, provided that the milk is sufficiently aerate and the consistency of the cheese is suitable for eye formation (my emphasis). On a proper assessment of the facts, therefore, I am not persuaded that the presence of eyes is an essential characteristic of Gouda. Consequently the appellants must fail on this aspect.
[13] The second point of difference between Walstra and Wessels raises problems of a different kind. Walstra appears to be correct in stating that the general consensus among cheese experts is that Gouda ripens for at least five weeks at temperatures over ten degrees Celsius. According to standard C-5, which is the standard for Gouda cheese prepared by the United Nations Food and Agricultural Organisation and the World Health Organisation, Gouda is ripened at a temperature of between 10 and 20 degrees Celsius for five weeks. (An exception is Baby Gouda which requires a two week maturation period only. The cheese under consideration in this appeal is not Baby Gouda and nothing further needs to be said in respect of this variety.) Ripening at a lower temperature tends to slow down the maturation process. I do not understand Wessels to disagree with Walstra’s views in relation to the ordinary ripening time and temperature required for the production of Gouda but it was her opinion that a cheese develops its taste gradually and a distinctive Gouda taste should be detectable after a period of approximate three weeks.
[14] The question that arises is when is a classification of cheese to be made by the Commissioner for the purposes of the Act. Counsel for the appellant submitted that the classification should relate to the time of importation and that changes in the character of the article that occur thereafter may not be taken into consideration. Consequently, and according to the argument, as the cheese in question had at the time of importation ripened for only three weeks at temperatures of between two to five degrees Celsius, it could not be classified as Gouda, even if further ripening at a higher temperature and for a longer period after importation might result in the cheese attaining the Gouda characteristics. It may be noted that after ripening for three weeks in the Lactos factory much of the cheese now in question was in cold storage at two degrees Celsius for some six to seven months before importation. This led Walstra to doubt whether the cheese could thereafter ever develop a Gouda flavour and consistency. On the contrary, he said, there was a significant risk of strong “off” flavours developing.

[15] Goods are liable for customs duty at the time of entry into the Republic for home consumption, which is deemed to be the time when the bill of entry is delivered to the controller of customs and excise concerned (see sections 39, 44(1) and 45(1) of the Act). In terms of s 47(1) duty is payable at the time of entry for home consumption. All of this presupposes a determination of the classification at the time of importation. To this extent the appellants’ submission is correct. It does not necessarily follow from this that the determination of a classification cannot thereafter be altered if subsequent events establish that the original classification was incorrect. Indeed in terms of s 47(9)(d) the Commissioner is entitled to amend or withdraw a determination and make a new determination with effect from the date of first entry of the goods or the date of the original determination or the date of the new or amended determination. In all events there is no reason in principle why cheese cannot be brought under a particular heading simply because the recommended ripening period has not yet elapsed. Parmesan cheese, for instance, is properly matured after a year but that is not to say that it is not to be regarded as Parmesan if it is imported before the full ripening period is completed. The maturation of cheese at the correct temperature and for the proper period no doubt affects the flavour and quality of the cheese but not its essential character. Walstra indicated that classification of cheese is possible only when its “eating quality” (i.e. its flavour and consistency) has developed sufficiently. This appeal is, however, concerned with a classification for tariff purposes under the Act and not with the cheese’s suitability for eating purposes. Of course it might be difficult to place cheese into a particular category before it has ripened properly. Moreover, classification at a very early stage might be impossible. In the present case the cheese had reach a relatively advanced stage of production. After the final removal of the whey, the curds were pressed into moulds and then immersed in a brine solution. After brining the cheese was removed from the solution, washed, dried and vacuum packed before being stored for three weeks at two to five degrees Celsius. At that stage classification, though difficult, was possible.
[16] The cheese in issue in this appeal was released to the corporation and, presumably, was sold in the market place. Whether it was sold as Gouda was not disclosed by the appellants. We are left to consider whether the cheese was Gouda at the time of importation and, for the reasons set out, the fact that it had not ripened in accordance with accepted Gouda standards does not in itself preclude it from being classified as Gouda, although it makes the task more difficult.
[17] Apart from the different percentages relating to FIDM content and, possibly, the type of mesophyllic starter culture used, the processes for manufacturing Gouda and Edam are not dissimilar. According to Walstra the type of mesophyllic starter culture used by Lactos did not produce carbon dioxide and, therefore, eyes. Significantly enough, he added:

“It is to be noted that in other respects the body of the cheese has a consistency which would lead to a cheese with eyes, provided a gas-forming starter culture is used. Moreover, in my view, many aspects of the manufacturing process described by Mr White resemble that used for Gouda. It is clear, however, that a key element - the appropriate starter culture - is missing. This has led to the absence of the tell-tale Gouda eyes.”

[18] Counsel for the appellants pointed to five differences between the Edam and Gouda processes in the Lactos factory. But the differences, according to Wessels, seem to have little practical significance. Walstra agreed that many aspects of the process used to manufacture the cheese in issue resembled that used for Gouda. His only qualification related to the starter culture used. According to White, however, the same mesophyllic starter cultures are used in his factory to produce both Gouda and Edam, the sole difference being that double the quantity is used for Gouda. The type of starter culture used - and according to Wessels there are a number that may be used for both Gouda and Edam - affects the consistency and texture of the cheese and its flavour. However there is no evidence to show what starter culture Lactos used or that it was of such a nature as to disqualify the cheese from being classified as Gouda on the grounds of consistency, texture or flavour. Walstra, according to the passage quoted in para 17 above, considered the starter culture to be inappropriate only because it did not produce eyes.
[19] The consistency of Gouda should be hard or semi-hard and suitable for slicing. Although Walstra complained about the “plastic consistency” of the Australian cheese, he added that its

“consistency might be regarded as being within the range for Gouda, although only maginally so because the cheese is on the soft side”.

Wessels said that the consistency of the cheese was “reconcilable” with the character of a “green” Gouda, i.e., a Gouda that had not yet fully matured.
[20] Both experts were of the view that the cheese had a bland or, possibly, flat taste. Wessels considered that it could nevertheless be identified as a “green” or “young” Gouda. Walstra was of the view that the natural cheese had “some flavour, which to some extent resembled Gouda” but that the orange-coloured cheese had very little flavour, “certainly nothing vaguely resembling that of a Gouda”.
[21] To sum up at this stage:
The cheese now in question was manufactured from milk normally used for the production of Gouda and the fat content of the cheese, expressed as a percentage of dry matter, fell within the Gouda range.
Moreover the process of manufacture of the cheese was identical to, or closely resembled, the process ordinarily used in the manufacture of Gouda.
The absence of eyes in the body of the cheese and the fact that it had not matured properly at the date of importation do not preclude it from being classified as Gouda for the purposes of the Act.
The consistency of the cheese was within the Gouda range. There was, however, disagreement between Walstra and Wessels on whether the flavour of the cheese - especially the orange-coloured cheese - was compatible with the Gouda flavour. There is no compelling reason why either expert’s view should prevail in this regard.
[22] On a proper assessment of the evidence, and apart from the dispute between Walstra and Wessels concerning the flavour of the cheese, there are no adequate grounds for setting aside the Commissioner’s decision. It only remains to decide how to resolve the disputed issue concerning the cheese’s flavour. Counsel for the Commissioner submitted that the onus rested on the appellant to show that the Commissioner’s determination was wrong. The submission was based on the erroneous assumption that the appellants had to discharge the onus that ordinarily rested on an applicant in motion proceedings. The fact of the matter is that the present proceedings are not motion proceedings in the usual sense. It is an appeal in terms of a statutory provision. It is arguable that the question of onus may be covered by s 47(9)(b) of the Act which provides that any determination made by the Commissioner shall,

“subject to appeal to the court, be deemed to be correct for the purposes of this Act, and any amount due in terms of any such determination shall remain payable as long as such determination remains in force”.

The matter was raised but left open in Commissioner for Customs and Excise v CI Caravans (Pty) Ltd 1993 (1) SA 138 (N) at 149 A-B. It is also unnecessary in this appeal to decide whether the deeming provision in s 47(9)(b) has relevance to the question of onus. It appears to me that the matter can and should properly be determined by applying proposition 9 in the oft-quoted decision of R v Dhlumayo and Another 1948 (2) SA 677 (A) at 706, viz, that where an appellate court is merely left in doubt as to the correctness of a trial court’s decision it will uphold it. The fact that the Commissioner does not exercise a judicial function does not affect this principle once it is accepted that an appeal under s 47(9)(e) is an appeal in the wide sense. Recognising as I do that Walstra’s views are not to be preferred above those of Wessels on any of the material points in issue, it follows that, putting it at its highest for the appellant, there is merely doubt as to the correctness of the Commissioner’s decision. The appeal, therefore, cannot succeed.
[23] The appeal is dismissed with costs including the costs consequent upon the employment of two counsel.

L S MELUNSKY AJA

CONCUR:
SCUTZ JA)
NUGENT AJA)