OPINION OF MR ADVOCATE-GENERAL MAYRAS

DELIVERED ON 5 OCTOBER 1972 (*1)

Mr President,

Members of the Court,

I — The facts

The preliminary questions submitted to the Court by the College van Beroep het Bedrijfsleven (the Netherlands Court of Appeal in economic matters) in accordance with Article 177 of the Treaty of Rome lead the Court to interpret certain provisions of secondary Community law. Consideration of these questions also requires the Court to make an incursion into the realms of organic chemistry.

This dispute in tact relates to tne application of certain methods of chemical analysis of a product intended for human consumption which was brought before the Netherlands court in circumstances of which I shall briefly remind the Court.

On 18 June 1970 the undertaking Handels-maatschappij Stern en Spierenburg, the subsidiary of the limited company Ver-eenigde Oliefabrieken of Rotterdam, declared for export to Bolivia a consignment of 100698 kg of refined lard packed in 6000 tins each weighing 16.680 kg.

In order to benefit from tne export refund provided for this product, the undertaking gave an assurance that the lard conformed to the standards of quality prescribed by the Community regulations. The refund, which is equal to the difference between the price on the world market and the price in the Community, may in fact be granted at the time of the exportation to a third country of lard and other rendered pig fats coming under tariff heading 15.01 A II in implementation of Regulation No 121/67 of the Council of 13 June 1967 on the common organization of the market in pigmeat. The general rules for granting export refunds and the criteria for fixing the amount of such refunds were laid down by Regulation No 177/67 of the Commission of 27 June 1967 whilst the Commission in Regulation No 2403/69 of 1 December 1969 laid down special conditions intended to restrict refunds to products complying with certain criteria as to quality with regard to their composition, preparation and packaging. Annex I requires, amongst other criteria, that the lard shall have a minimum ‘Bömer’ value of 73 and stipulates that controls for the observance of this condition shall be carried out according to the method known as the diethyl ether method or the acetone method. To describe this method the provision refers to a draft recommendation of the IOS (International Organization for Standardization) concerning the sampling and analysis of animal fats.

A first analysis was carried out according to the method of the Netherlands Pharmacopoeia by the Rijkstoezicht op de Bereiding van Eetbaar Vet (National Office for the Supervision of the Preparation of Edible Fats) before the lard was packed, on an average sample of the entire consignment; a ‘Bömer’ value of 74.3 was found, thereby exceeding the value prescribed by the Community Drovisions.

However, at tne time of the conclusion ot the customs formalities, the Netherlands customs authority took two of the tins of lard and had them analysed by the Rijkszuivelstation (National Institute for Dairy Products). This control determined the ‘Bömer’ value at 72.2 and 72.5 respectively for each of the samples taken. On the basis of this analysis, from which it was clear that the lard exported did not reach the minimum value required, the Produktschap voor Margarine, Vetten en Oliën (Public Agency for Margarine, Fats and Oils) refused to pay NV Vereenigde Oliefabrieken the amount of the refund for which it had applied.

When the matter was brought before the College van Beroep voor het Bedrijfsleven it decided to stay proceedings until the Court could give a ruling on the following preliminary questions:

1.Must Article 2 of Regulation No 2403/69, read in conjunction with Article 1 of Regulation No 1041/67 and any other relevant provisions of Community law, be interpreted to mean that the control referred to by this article must be carried out exclusively on samples taken on the day laid down in Article 1 of Regulation No 1041/67 or must Article 2 be interpreted to mean that such control may also be carried out on samples taken some time before the said day?

2.Must Article 2(1) of Regulation No 2403/69 and Annex 11(1) to that regulation, read together with the footnote to that annex, be interpreted to mean that the ‘Bömer’ value of refined lard must be determined according to the method laid down in Annex 11(1) of Regulation No 2403/69 in the following manner: ISO/TC 34/SC 6/WG 3/N 73?

3.Does a correct interpretation of Article 2(1) of of Regulation No 2403/69 imply the restriction of the control by sampling, understood in the sense envisaged by this provision, of the ‘Bömer’ value of a quantity of refined lard having a nett weight of 100698 kg packed in 6000 tins to the analysis of the contents of two of such tins?

4.Must Article 2(1) of Regulation No 2403/69 be interpreted to mean that only the result of an analysis based on a control by sampling, understood in the sense envisaged by this provision, must be taken into consideration in deciding whether the condition as to the ‘Bömer’ value has been fulfilled or, on a correct interpretation of this provision, must account be taken in addition, or if appropriate exclusively, of the result of an analysis fulfilling the conditions laid down in this respect by the same provision and adopted thereunder, despite the fact that such analysis was not carried out on samples taken by random sampling but on the basis of a large number of samples taken continuously from the relevant quantity?

II — The time when the samples must be taken for the control prescribed by Regulation No 2403/69

The first question submitted thus relates to the time when the control prescribed by Regulation No 2403/69 must be carried out. Nevertheless, although Annex I to this provision lists the conditions as to quality which the exported lard must fulfil, requiring in particular that this product should have a minimum ‘Bömer’ value of 73, and although Article 2 of the regulation stipulates that the control by sampling of the conditions laid down in Article 1 shall consist of an organoleptic examination and physical and chemical analyses carried out in accordance with the methods referred to in Annex II, no provision in this regulation expressly fixes the time when the samples must be taken for the control. NV Vereenigde Oliefabrieken, the plaintiff in the main action, infers from this that the sample may be taken at a date prior to exportation. It argues that lard is a stable product the ‘Bömer’ value of which remains invariable, and that consequently, a control of this value carried out some time before exportation is valid and fulfils the requirements of Regulation No 2403/69, at least, if the sample is correctly taken and is representative of the consignment as a whole.

The Netherlands Government takes the same view. It considers that in the absence of any binding provision in the Community regulations a distinction must be drawn between perishable foodstuffs, which must be sampled and analysed at a date as close as possible to that of the customs export formalities, and products the characteristics of which are stable — as is the case with lard — where no disadvantage is occasioned if this control takes place at an earlier date.

Although the Commission recognizes that its Regulation of 1969 is silent on this point, it does not share this view. It maintains that other provisions must be taken into consideration and that they imply that this control should be at the same time as the conclusion of the customs formalities. In this connexion, it refers to its Regulation No 1041/67 of 21 December 1967, which lays down detailed rules for the application of export refunds on products subject to a single price system including pigmeat and pig fat. Article 1 of this regulation provides that for the purposes of determining the rate of the refund — at least where the refund is not fixed in advance and in any event in order to determine any adjustments to be made to it — the operative date shall be the day on which the customs authority accepts the declaration of the exporter, that is to say, the document setting out his intention to export the product in question and to qualify for a refund.

From the point when the authority has indicated its acceptance, which, for the purposes of the regulation, is considered as the completion of the customs formalities, the products are placed under the control of the customs until they leave the geographical territory of the Community.

Article 1 (3) provides that the day on which the customs formalities are completed shall be the operative date for determining not only the quantity but also the nature and characteristics of the product exported. Thus taking these provisions together with those of Regulation No 2403/69 which sets out in detail the characteristics required of certain pigmeat products, especially of lard, it must be inferred that control of those characteristics, or at least of the taking of samples, must be carried out at the same date as the completion of the customs formalities.

This argument is based first of all on the same passage of Regulation No 1041/67 which is expressly referred to by the 1969 Regulation, Article 1 of which provides (here I shall quote it): ‘Without prejudice to other Community provisions, and in particular those of Regulation No 1041/67’ and requires exporters to declare at the time of the conclusion of the customs formalities that the products in question fulfil the conditions as to quality listed in Annex I to Regulation No 2403/69.

However, in my view, this line of argument also corresponds to the self-same objectives which the Commission wished to attain in requiring a control of the quality of products exported from the Common Market. Such control in fact forms part of the system of refunds intended to favour exports to third countries. It must allow both national administrations and the Community services to ascertain that the product which qualifies for a refund upon exportation is in fact the kind of product defined by the Community regulations and its heading in the Common Customs Tariff; the exact identification of the product is certainly a basic requirement of this system.

Secondly, control of the characteristics of the product intended for exportation must be carried out in such a manner and at such a time that the control is uniform, has the same value and credibility whatever the country of origin of the product and whatever the differences which may exist between the national rules.

Finally, the tests and analyses prescribed must be carried out on samples taken on a date which enables the characteristics of the product to be ascertained at the very point when exportation is about to take place. This is clearly imperative with regard to perishable foodstuffs. Nevertheless, even if it is conceded that all or part of the physical and chemical characteristics of certain products do not vary, or vary only very slightly with the passage of time, as the Court has been informed is the case with lard provided it has been perfectly packed, it would be both difficult and undoubtedly rash to make exceptions regarding such products, to the general rule thereby permitting control of their characteristics to be carried out at a date before that of completion of the customs formalities.

The product must thus be identified with caution as, even without envisaging an act of fraud, can one exclude the possibility of a simple error in the treatment of the product between the stage of manufacture and the stage of packaging or deterioration of the product as a result of faulty packaging? Likewise, if it is conceded that sampling and analysis for the purposes of control on exportation may be effected in the factory, what is the appropriate period within which the goods must actually be exported for such analysis to be considered valid? This would give rise to a wide variety of different cases which the administrations and national courts could no doubt appraise individually, but would not this involve abandoning the minimum degree of uniformity and coherence required by the application of a Community provision, which was recognized as necessary by the Court when it stated that ‘The common organizations of the agricultural market … can only fulfil their functions if the provisions to which they give rise are applied in a uniform manner in all the Member States’ (Judgment of 18 June 1970, Case 74/69, Hauptzollamt Bremen-Freihafen v Waren-lmport-Gesellschaft Krohn and Co. ([1970] ECR 451, at 459) and likewise: Judgment of 6 June 1972, Case 94/71, Schlüter & Maack v Hauptzollamt Hamburg-Jonas)!

Reading together the two regulations which the Court is requested to interpret, on the one hand, and the practical requirements of certain identification of the product, from an identical procedure of control no matter from what country the goods are exported, on the other hand, and finally the requirement that analysis of the characteristics which the product actually displays should occur when it is placed under customs control thus require that the first question should be answered to the effect that the taking of the samples in order to carry out the control by sampling must coincide with the date of completion of the customs formalities within the meaning of Regulation No 1041/67.

Ranged against tnis view, l perceive only one argument and that does not seem to me decisive. To qualify for a refund the product exported must meet stringent requirements as to quality. Taking into account the cost price, exportation is only rendered profitable by this refund the rate of which amounts in the present case, according to the plaintiff in the main action, to 25 % of the cost of the goods. It is thus necessary for the exporter to know in advance whether or not he can count on the refund. In other words, he must be assured at the stage of manufacture or in any case at the stage of packaging that his product indeed displays the characteristics required by the Community provisions. It is therefore necessary for him to know the results of the tests and analyses before submitting his export declaration to the customs authorities. If the sample is only taken on the day when the goods are placed under customs control and if in addition the results of the analysis can in practice be learned only some time afterwards (in the case brought before the Netherlands court we know that although the sample was taken on 18 June 1970 the first result of the analysis was only notified to the producer on 24 July, that is, two weeks after the lard left Netherlands territory for Bolivia) the exporter runs a serious risk, if this control establishes that the product does not fulfil the criteria laid down.

NV Vereenigde Oliefabrieken indeed relies upon the results of the analysis of an average sample of the consignment of lard in question, taken in the course of manufacture in accordance with the Netherlands provisions on health control of products intended for human consumption.

However, is it further necessary that such analysis, in order to be valid under the system of refunds, should be carried out according to the method not merely proposed but imposed by the Commission, as the Court knows, and not in accordance with the method of the Netherlands Pharmacopoeia? Indeed the latter method is also based on the use of diethyl ether but the Court has not been informed whether it is completely identical with that prescribed by Regulation No 2403/69.

The multiplication of controls of this nature, some national others Community, is without doubt undesirable. It is certainly desirable that they should be coordinated. Be that as it may, I consider that it is the duty of the exporter to make every endeavour in this sphere and to take all appropriate precautions to ensure that the analysis of his product before it is submitted to the customs authority conforms as strictly as possible to the analysis to which it will be subjected by the customs authority in checking the entitlement to the refund.

The second question submitted relates to the method of analysis which must be employed in order to determine the “Bömer” value of refined lard within the meaning of Regulation No 2403/69 and Annex II thereto.

It is not disputed that this value, which enables the purity of lard to be checked, must reach a minimum value of 73 in order that Community lard may qualify for the refund. As the Court was told in the oral procedure, completely pure lard attains a value of 75. On the other hand, according to a letter from NV Vereenigde Oliefabrieken to the customs authority of Rotterdam a value of only 71 is sufficient to show that the product analysed is in fact composed of pig fat. A certain margin of tolerance was taken into account in fixing the value at 73; according to the Commission, below this minimum level the product analysed contains other animal fats such as fats of bovine cattle, horses or sheep. In that case exportation of the product—which must, moreover, be classified under heading 15.06 of the Common Customs Tariff—does not confer entitlement to a refund.

This means that determination of the “Bömer” value is of decisive importance in implementing the system of refunds. It also means that in order to avoid as far as humanly possible any disparity of treatment between exporters it seems necessary at Community level to choose a specific method of analysis. This is the decisive point underlying the question submitted by the Netherlands court.

Is it necessary to prescribe as the only method applicable that referred to in Annex II to Regulation No 2403/69, that is to say, the method which forms the subject-matter of a draft recommendation of the IOS, or has the Commission merely advocated this method without precluding recourse to other methods more or less approximate in their procedure and degree of accuracy?

The plaintiff in the main action maintains that although Annex II prescribes only the method known as the diethyl ether method or the acetone method it mentions the draft recommendation of the IOS simply as a “source”. From this the plaintiff infers that this method, which has not yet been officially adopted by the International Organization does not preclude certain other methods, provided at least that they are based on the use of ether or acetone. This is true in the case of the method employed in accordance with the Netherlands Pharmacopoeia on the basis of which a first control was carried out in the course of manufacture on the quality of the lard in dispute.

This line of argument is based on a hasty, and in my view false, interpretation of the word “source” which NV Vereenigde Oliefabrieken only considers as having a purely indicative value and does not have the status of a binding provision.

On the other hand, the Government of the Netherlands concurs on this point with the Commission and recognizes that the IOS method is indeed the only one recognized by the Community provisions for the valid determination of the “Bömer” value of refined lard.

I also think that the use of the word “source” may easily be accounted for by the fact that, although the IOS has not yet adopted a final and officially applicable recommendation, the Commission made a clear reference to the draft prepared by that institution which was nevertheless known as such and identifiable by everyone through the references indicated in Annex II to the Community regulation.

By making an express reference to this draft the Commission adopted it, conferring on it legal force in Community law, so that it is to a certain extent integrated into the regulation itself. The Commission could indeed have reproduced the exact text of the draft recommendation and could equally well have referred to another method described in a technical treatise. Nevertheless, it is of little importance that the IOS had not and even today has not yet finally adopted this draft. It has become part of a regulation directly applicable in the Member States of the Community, as is stated moreover, in the final provision of the regulation.

Furthermore, in order to avoid any doubt should the IOS amend its draft, the Commission was careful to specify in a footnote that the methods of analysis referred to in Annex II, particularly the method of determining the “Bömer” value “are those operative on the date of the entry into force of this Regulation”, that is to say, for the products coming under tariff heading 15.01 A II on 1 January 1970.

The interpretation which I thus wish to suggest that the Court should adopt leaves even less doubt—perhaps more than with regard to the time when the samples must be taken—that the need for a uniformly applicable rule fully justifies the choice by the Commission of one and the same method of analysis for determining the “Bömer” value which is difficult to calculate, as we have been informed, and the results of which depend to a not inconsiderable extent on the method employed. I should add that it is further probable that the results also vary in a strict ratio to the skill of the chemist and to the manner in which the method is applied in practice although it appears to me that the description of the process was clearly explained in sufficient detail by the IOS document.

It is this possibility of a distortion, even a slight one, in the results of an analysis carried out on samples having the same origin which in all probability explains the slight differences found between the first analysis carried out at the request of the customs authorities by the Rijkszuivel-station of Leiden and the second analysis carried out on 28 August 1970 on a sample from the same tins of lard.

In the former instance the control gave a “Bömer” value of 72.2 and 72.5, that is an average of 72.35.

In the second instance the average figure obtained was slightly higher: 72.5 without however, attaining the minimum necessary of 73.

The fact that two successive controls separated by some weeks and carried out by the same laboratory on samples taken from the same packages produced results differing so slightly is a sufficient indication of the margin of relative uncertainty which can scarcely be imputed to anything other than human causes. Nevertheless, in my view this constitutes a additional argument in favour of selecting a single method of analysis. By this means at least one might hope to eliminate the differences inevitably caused by the use of different methods. In this way an explanation may likewise be found for the more significant differences found between the results of the first analysis carried out on an average sample taken in the course of manufacture by the Rijkstoezicht op de Bereiding van Eetbaar Vet and the results of subsequent analyses carried out at the request of the customs authorities.

I advocate in the strongest terms that in this matter the Court should reply to the second question submitted to the effect that the method known as the diethyl ether method or the acetone method recommended by the IOS at the reference indicated in Annex II to Regulation No 2403/69 excludes all other methods, is obligatory and is the sole valid method for the purposes of the control by sampling of the “Bömer” value of lard required by this regulation.

It remains to consider two questions which, it must indeed be recognized, are less concerned with calling in question the interpretation of Community law than with the appraisal of fact bound up with the particular circumstances of the case brought before the College van Beroep voor het Bedrijfsleven. In the reply which the Court gives to the first question, that court expects to obtain a definition of what is intended by “control by sampling”. None of the provisions of Regulation No 2403/69 gives the slightest indication as to what should be understood by “sampling” and, with regard to the determination of the “Bömer” value, the IOS method merely explains that the sample of lard to be analysed must weigh approximately 100 gr without indicating the relative importance of the sample in relation to the quantity of the product controlled. Moreover, since the court making the reference is fully aware that what is intended by “control by sampling” may differ greatly according to the nature of the products, it can only ask the Court to give its interpretation in the particular circumstances of the case, that is: is it lawful to restrict such a control to the analysis of the contents of two tins of lard out of a quantity of 6000 in order to determine the “Bömer” value of the consignment as a whole?

I am tempted to urge the Court to reply that this is in fact within the scope of the appraisal of the national court and that it is not appropriate for this Court to participate in the consideration of a problem the solution of which is all too clearly linked with the particular facts of the case brought before the national court. Nevertheless, it seems to me necessary to endeavour to provide it with a guideline whilst refraining from any claim to deliver a general interpretation valid for all cases. It seems to me that the reply of the Court must be formulated so as to take into account the nature of the relevant product and the form in which it is exported and thus its packaging.

If the product in question is relatively dissociable, displays physical and chemical characteristics which are not particularly stable and which is transported if not in bulk at least in containers of large capacity, I am led to consider that the taking of samples must of necessity be carried out at different points of the total mass of the product and must relate to a relatively considerable portion of such mass so as to ensure that the consignment is homogeneous and that the samples taken are sufficiently representative of the whole.

In my opinion, the situation differs when the product is packed in metal containers having a very small capacity in relation to the total amount exported, is thus ready for sale and, in addition, has stable characteristics and is not in danger of dissociating.

When this is the case, as it is for a large number of foodstuffs, and when, as in the present proceedings, the exporter himself declares to the customs authorities that the contents of each of the packages come from one and the same manufacturing run, in my view the taking of a few samples or indeed of one should be regarded as according with the concept of “control by sampling”. Indeed, it is impossible to understand on what grounds a particular tin of lard is more representative of the consignment as a whole than any other of the tins which make up the consignment. I should like to add that if the exporter were to have any reason to dispute a control, considering that it related only to an insufficient quantity of samples, he could request this to be checked by another analysis with more varied samples, provided always that he is entitled under the relevant national legal system to demand such a review, as is the case in the Netherlands.

I therefore wish to suggest that the Court should reply to the questions submitted to the effect that, since the product in question is stable and has been packed in such a way to retain its characteristics, and since the exporter has declared that the product is of uniform quality, the expression “control by sampling” should be interpreted as applying validly to the contents of a number, even of a very limited number, of packages.

The last question submitted to the Court by the Netherlands court appears to call for a reply containing similar subtle distinctions. It relates to the nature of the sample analysed and asks the Court whether, for the purposes of Regulation No 2403/69, account may be taken of an analysis effected on an average of a large number of samples taken continuously. In this respect, too, the question is inseparable from the circumstances of the case. The continuous taking of multiple samples, that is to say, throughout a manufacturing process, does not in principle fulfil the requirements of the Community provisions if only because of the rule the existence of which I have admitted with regard to the time when the samples must be taken, unless the factory itself is to be placed under customs control, a case clearly restricted to the realms of theory.

When a product of a highly homogeneous nature is exported in packaging, it is difficult to understand the point of taking a large number of samples in order to establish an average sample. This brings us back to the preceding question for, when the exporter guarantees that the consignment submitted to the customs authority is of uniform quality—and this carries even more weight since the goods in question are sold under a commercial mark—the determination of an average sample is quite unnecessary.

In any case, no provision, either in the relevant regulation or in the technical annexes thereto, explains the method of taking the samples and I consider that this problem must be left to the appraisal of the national authorities subject to review by their courts. If the Commission did not consider it worthwhile to provide more thorough and detailed legislation in this respect it is because it intended to leave this task to the competent authorities of the Member States or, at least, to leave them to resolve the actual problems of carrying out controls on exportation.

Finally, since this is an essentially technical field, it is certainly undesirable, and furthermore I do not consider it possible, that the Court should on the application of national courts put itself in the stead of the Community authorities which have the power to adopt regulations in this field.

I am thus of the opinion that the Court should reply to the first two questions submitted by the College van Beroep voor het Bedrijfsleven as I have stated above; I suggest that the Court should provide answers to the other two questions which, without laying down any general principle of interpretation, take into account the facts peculiar to the nature, characteristics and packaging of the product in question, at the same time stating that the national courts enjoy a wide power of appraisal so that the solution of questions of this nature may be adapted to the particular circumstances of each case.

*

(1) Translated from the French.