1. 

The case to be decided by the Court is concerned principally with two problems, both of great constitutional importance: the first concerns the review which the Court of Justice may effect — in direct actions — of measures of the Parliament; it is then necessary to establish what powers, if any, the Parliament has in deciding where its offices are to be situated and where it is to work, having regard to the agreements hitherto arrived at in this matter by the Member States. The Court has already had occasion to consider the review of measures of the Parliament, in a restricted and specific context: I refer to the recent judgment of the Court in Lord Bruce of Donington v Aspden. It is, however, the first time, I think, that the Court has been requested to rule directly on this matter. On the other hand, no case-law exists on the agreements concerning the seat of the Parliament.

That stated, I shall now give a brief summary of the facts of the case.

By a resolution adopted on 7 July 1981 the European Parliament set out its views on the seat of the Community institutions and, in particular, on its own seat. Pending a decision on a single location for its work, it took a series of decisions, in particular:

By a document lodged at the Registry on 7 August 1981 the Grand Duchy of Luxembourg challenged the abovementioned resolution, claiming that it should be declared void on grounds of lack of competence and infringement of essential procedural requirements. In its defence the Parliament contended that the application was inadmissible on a number of grounds and, in the alternative, that it was unfounded.

2. 

In this opinion I shall deal first of all with the numerous grounds of inadmissibility which have been advanced by the defendant. These grounds constitute objections which always, or almost always, involve complicated points of law. I shall then proceed to consider the substance of the dispute, that is to say, the two grounds of lack of competence and infringement of essential procedural requirements pleaded by the applicant. In this context I must above all review, in chronological order, the contents of the various agreements arrived at by the Member States with regard to location of the Parliament's place of work.

3. 

I shall begin by considering the procedural factors.

The defendant institution argues on various grounds that the application is inadmissible and, for ease of treatment, I shall assemble these grounds in three groups. The objections in the first group concern the status of the applicant, those in the second group are based on the characteristics of the contested measure, whilst those in the third group are founded on the definition, on the basis of Article 38 of the ECSC Treaty, of the measures of the Parliament which may be directly contested.

I propose to analyse the admissibility of the application in accordance with that division and I shall consider first of all the objections concerning the Grand Duchy of Luxembourg's title to sue. The first objection of this nature proceeds from the fact that the three Treaties reserve power to fix the seat of the institutions to the governments of the Member States. Article 77 of the ECSC Treaty provides that “the seat of the institutions of the Community will be determined by common accord of the Governments of the Member States”; similarly Articles 216 of the EEC Treaty and 189 of the EAEC Treaty lay down, both employing the same wording, that “the seat of the institutions of the Community shall be determined by common accord of the Governments of the Member States”. According to the Parliament, the foregoing requires the Member States to take all the steps necessary to reach an agreement. The fact that one of the States, whilst having available to it procedures such as negotiation and agreement to settle the problem of the seat, resorts to court proceedings makes the application inadmissible. The situation gives rise to a particular form of bar to action based on the attitude and conduct of the applicant. That bar is founded on the rule of estoppel which prevails in ordinary international law and is incorporated into the Community system by Article 5 of the EEC Treaty.

The objection appears to me unfounded. Estoppel is a characteristic feature of Common Law systems and its purpose is to prohibit a person from performing certain acts where those acts are at variance with the result partly brought about by that person's conduct. That principle, or one similar to it, is employed in international law to prevent a State from relying upon rights based on failure to implement an agreement when, although it was capable of implementing the agreement, it did not make any effort to do so (cf. the judgment of the International Court of Justice of 20 February 1969, Federal Republic of Germany v Denmark, Federal Republic of Gennany v Netherlands, ICJ Reports 1969, p. 26, at para. 30).

Nevertheless the rule of estoppel, the limits to which are already very vague in relations between States, may not be transposed into Community law and in any event is not applicable to this case.

It must first of all be borne in mind that the Community system is endowed with an organic structure and has a judicial body to settle disputes between Member States and between such States and the institutions. It is thus reasonable in such a system that the persons concerned should resort to court proceedings in order to protect their respective rights. It would be absurd to consider in breach of good faith an application to the Court, which is required to “declare the law”, when the existence of a scat for the institutions forms one of the salient features of the Community order. This point of view is confirmed by the case-law of the Court of Justice. I refeito the judgment of 31 March 1971 in Case 22/70 (Commission v Council [1971] ECR 263). Amongst the points at issue in that case was whether the Commission had lost its title to sue because it was itself “responsible for the situation in question through having failed to take, at the proper time, the steps necessary to allow Community powers to be exercised, by submitting suitable proposals to the Council” (paragraph 62). The Court declared that since the case was concerned “with the institutional structure of the Community, the admissibility of the application [could] not depend on prior omissions or errors on the part of the applicant” (paragraph 63).

Furthermore, it does not appear to me that the rule of estoppel is referred to in Article 5 of the EEC Treaty. In so far as that provision is relevant it merely provides that: “Member States shall take all appropriate measures, whether general or particular, to ensure fulfilment of the obligations arising out of this Treaty... They shall facilitate the achievement of the Community's tasks.” In this case the Member States have not arrived at a final agreement on the location of the seat of the Parliament; that, however, does not imply any failure to fulfil an obligation arising out of the Treaty for the simple reason that no such obligation exists. Articles 77 of the ECSC Treaty, 216 of the EEC Treaty and 189 of the EAEC Treaty do not require the Member States to conclude an agreement on the seats of the institutions but merely reserve to them the power to do so, provided, of course, that their failure to exercise that power, or their wrongful exercise thereof, does not lead to the paralysis of the institutions. That much emerges from the letter of the said provisions and from the circumstances that, according to the prevailing view in ordinary international law, pacta de contrahendo give rise, at most, to a duty to conduct negotiations concerning the future agreement in good faith.

As an example, provision is made for such a case in Article 220 of the EEC Treaty, which provides that:

“Member States shall, so far as is necessary, enter into negotiations with each other with a view to securing for the benefit of their nationals :

The protection of persons and the enjoyment and protection of rights under the same conditions as those accorded by each State to its own nationals;

The abolition of double taxation...;

The mutual recognition of companies or firms...;

The simplification of formalities governing the reciprocal recognition and enforcement of judgments of courts or tribunals and of arbitration awards.”

It will be noted that that provision, the purpose of which is to emphasize the duty of the Member States to negotiate, uses wording (“enter into negotiations”) which is much more precise than that used in Articles 77 of the ECSC Treaty, 216 of the EEC Treaty and 189 of the EAEC Treaty (“the seat... will be determined by common accord...”; “the seat... shall be determined by common accord...”). Furthermore the first provision set out, in general terms at any rate, the content of the future agreements. If therefore not even a duty to negotiate can be clearly discerned in the provisions of the Treaties on the seats of the institutions, it appears to me difficult to maintain that the Member States have infringed Article 5 of the Treaty in this respect or that their inertia has impaired the working of the institutions. In fact it is established that the Member States have broached the problem of the seat of the Parliament on a number of occasions, concluding agreements whose content and legal status I shall deal with in considering the substance of the dispute.

Article 5 furthermore requires the Member States to facilitate the achievement of the Community's tasks: but even here I do not see how it may be considered that those States have not fulfilled their obligations by failing to attain final agreement on the seat of the Parliament. In spite of all the limitations the agreements to which I have referred show that the Member States have complied with that requirement.

4. 

The second objection also relates to the status of the applicant, but is based on a different argument: the various Member States, it is said, may not individually challenge Community measures which infringe the right conferred upon them by the basic sources to decide the seat of the institutions. The right to challenge such measures belongs jointly to all the Member States and accordingly they may only exercise it jointly. In this case the Grand Duchy is the only State which has brought an action: it follows that this application must be considered inadmissible for want of title to sue.

I cannot concur with that argument. I have already pointed out that the three Treaties, by uniform rules, entrust the Member States with power to determine the seat of the institutions. The characteristic feature of these provisions resides in the fact that in this particular field, which is of importance in constitutional matters because it is connected with the functioning of the institutions, the Community institutions are expressly deprived of power, which continues to fall within the sovereignty of the Member States. I emphasize the word “sovereignty”: why therefore should not the title of every Member State to take proceedings individually in defence of its own rights be recognized? Nor may a different conclusion be inferred from the fact that the choice of the seats requires the agreement of all the Member States: in fact the framework in which that exclusive power is exercised — a framework which presupposes such agreement — is one matter; quite another is the right of each Member State to act individually to counter any infringement of the reservation of power in question.

The two distinct aspects of the matter which I have described must not be confused. Any other view would result in an incongruity: the refusal of a single State to participate in a joint action would prevent all the others from exercising their right to defend before the Court their specific interest in settling the problem of the seats even though they are all agreed on that course. The view may thus be taken that each State has its own material interest in this matter which is recognized by the abovementioned parallel provisions of the three Treaties and it is accordingly beyond dispute that it has a title to institute proceedings in defence of that interest.

In this case the Grand Duchy of Luxembourg not only claims that the resolution of the Parliament is vitiated by lack of competence because it is contrary to the provisions of the Treaties whereby power is reserved to the Member States to decide the seat of the institutions: it also maintains that that resolution, by reason of its particular content, is contrary to the agreements hitherto concluded by the Member States on the provisional working places of the Parliament. As will subsequently emerge more clearly, the substance of the dispute revolves essentially around the question whether the decision of the Parliament falls within the framework of the agreements already concluded by the Member States on its seat without modifying them in any way, in which case it must be regarded as a measure by the Parliament regulating its internal organization, or whether it exceeds or is contrary to those agreements, in which case it is vitiated by lack of competence. The agreements hitherto arrived at with regard to the seat of the Parliament are of particular concern to Luxembourg. It is sufficient to have regard to the “Decision of the Representatives of the Governments of the Member States on the provisional location of certain Institutions and Departments of the Communities” of 8 April 1965: as is stated in the preamble, the decision was adopted precisely “in order to settle certain problems peculiar to the Grand Duchy of Luxembourg”, and with that in view it provides in Article 4 that: “The General Secretariat of the Assembly and its departments shall remain in Luxembourg.” That means that in this case Luxembourg has a specific interest in challenging the resolution of the Parliament, an interest which cannot therefore be treated as equivalent to the interest of the other Member States.

The Parliament disputes that proposition. It maintains that in fact the Member States do not each have an individual interest in settling the question of the seat through agreement amongst the governments but together have a single and indivisible interest. It follows that the Member States do not individually have the right to protect that interest before the Court. I have already discounted the view that the Member States have a single interest. Nevertheless, let us accept that view for the sake of argument. It would not prevent an individual Member State from seeking to defend its rights by legal action. It is in fact a generally recognized principle that the co-owners of a right may also act individually: to provide examples drawn from private law reference may be made to the position of persons jointly and severally liable and of the owners of common property. Furthermore, that principle meets the basic requirement that a remedy in the form of an application to the courts must be provided: to concede that in proceedings like the present case an application to the court might be made only by all the co-owners, acting together, of the right adversely affected would endanger the exercise of the right in good time in particular cases or render it entirely impracticable.

5. 

In the same line of thought the Parliament puts forward a third argument. It assumes for the sake of argument that a Member State may institute court proceedings alone, although it is only a co-owner of the right which it is defending: in order that the Court may give judgment it is nevertheless indispensable that all the other co-owners of the same right should be joined to the proceedings. That is to say it constitutes a case in which a party must necessarily be joined to the proceedings by virtue of the nature of the matter in dispute. The Parliament goes on to argue that the rules governing procedure before the Court of Justice do not provide for joinder by order of the court but only for voluntary intervention and that, as will be seen, is subject to certain limitations. Accordingly, those rules do not contain the means of ensuring that all parties are joined. As a result of that lacuna, whenever the matter before the Court involves necessary joinder and all the parties concerned in the matter are not before the Court, the latter cannot give judgment properly, that is to say, in such a way as to satisfy the applicant's claim. The conclusion to be drawn in this case is clear: the Grand Duchy must fail in its action.

This argument is no more persuasive than the earlier ones. I say so not only because it proceeds from the erroneous assumption that in the relevant field each Member State does not have an independent interest. It has a further and more specific weakness: the consequences which the Parliament represents as coherent and logical cannot in fact be inferred from the rules on intervention.

I shall begin by observing that there are two conditions for necessary joinder: that a number of persons must be concerned with the matter at issue and that a judgment given against one or a number of them would be nugatory. In this case there can be no doubt that if the Court of Justice ruled in favour of the Grand Duchy the judgment would not be nugatory: it would in fact prevent the Parliament from taking further steps in the matter and thus protect the interests of the applicant. Thus it does not seem to me proper to talk of “necessary joinder” in this case. It follows from this that it becomes irrelevant to establish whether or not intervention by order of the Court is a feature of the Community's legal process. In this respect it suffices to refeito the judgment of 10 December 1969 in Case 12/69 (Wonnerth v Commission [1969] ECR 577): such a form of intervention, the Court ruled, “is not provided for in the Rules of Procedure”.

The position of the co-owners of the right who have not instituted proceedings is in fact protected, as regards the action brought by the applicant, through the procedure of voluntary intervention and the procedure which allows third parties to challenge a judgment of the Court. I will remind the Court that in accordance with the rules in force notice is given of the institution of proceedings by publication in the Official Journal of the European Communities of “notice... of the date of registration of an application originating proceedings, the names and permanent residences of the parties, the subject-matter of the dispute, the claims made in the application and a summary of the contentions and of the main arguments adduced in support” (Article 16 (6) of the Rules of Procedure). All persons concerned are thus enabled to appraise, upon receipt of the essential information, whether intervention in the proceedings is appropriate or not.

The Parliament does, it is true, observe in this connection that the Rules of Procedure in force permit only intervention in support of one of the original parties, whereas in this case the other Member States might have intervened in order to advance arguments other than those of the applicant and not necessarily in harmony with those of the defendant institution. From this point of view, then, the protection which the system provides for persons having a common interest which is defended by one of them alone is relatively slight: if the arguments advanced by the applicant and by the persons having a common interest do not coincide the only remedy available to the latter is to challenge the judgment in third-party proceedings. This analysis of intervention is based on the provisions devoted thereto in the three Protocols on the Statute of the Court of Justice. The second paragraph of Article 34 of the Protocol on the Statute of the Court of Justice of the ECSC provides that: “Submissions made in an application to intervene shall be limited to supporting or requesting the rejection of the submissions of one of the parties.” The third paragraph of Article 37 of the EEC Protocol and the third paragraph of Article 38 of the EAEC Protocol provide in identical words that: “Submissions made in an application to intervene shall be limited to supporting the submissions of one of the parties.” The use of the words “limited to”, which recur in all three articles, seems to me to form an obstacle which is difficult to overcome for proponents of the view that forms of intervention other than intervention exclusively in support of one of the parties are permitted.

The point that the protection of persons not empowered to intervene independently is flimsy and insufficient is thus correct. But to proceed thence to accept that the co-owner of a right may not act alone is a difficult step to take and, in my view, impossible. In both cases there is a risk that the right to defend one's interests will be prejudiced. Of the two risks — the impossibility of applying to the Court other than as joint parties (that is, together with all the other proprietors of the right at issue) and the limitations placed upon the right of intervention, which are, however, offset by the right to challenge the judgment in subsequent third-party proceedings — the former appears to be much more serious. In that state of affairs, I consider that the interpretation of the system to be favoured is that which provides the fullest protection or, perhaps more accurately, the least incomplete protection, for all the interests involved. Accordingly, even if we accept the Parliament's argument founded on the idea that in this case the right at issue is single and indivisible, the conclusion must be that each State may act individually; however, the other parties having an interest may intervene in the proceedings at least in support of one of the parties, or they may, if appropriate, challenge the judgment as third parties.

6. 

The Parliament next argues on a second ground that the applicant has no title to sue. It avers that Luxembourg was not entitled to take proceedings to protect the interests of the staff of the Parliament and that the application is therefore inadmissible. This objection — which, if I may say, is somewhat peculiar — is based on the fact that the applicant, in challenging the resolution of the Parliament, has submitted inter alia that it adversely affects the officials of that institution. It is clear that the defendant's argument is entirely unfounded. The fact is that Luxembourg has never asserted that it is taking proceedings in the interests of third parties : on the contrary, it has always declared that it has taken proceedings in exercise of its own right, as should be clear from what I stated earlier concerning title to sue. I accordingly refer to what was said on that point.

7. 

I shall now proceed to consider the objections of inadmissibility based on the characteristics which a Community measure must display if it is to be open to challenge. The objections in question are that the resolution is in the nature of a confirmation and that it is only of internal effect.

The first objection can be summarized easily. The Parliament contends that the resolution of 7 July 1981 merely confirms the decisions taken concerning the places of work by the resolution adopted on 13 March of that year: it follows that the former may not be challenged independently, for it was merely a confirmation, whilst the latter measure, that of 13 March, was not contested in good time. In this connection it should be remembered that an application against measures of the Parliament must be submitted “within one month of the publication of the act of the Assembly” (see the second paragraph of Article 38 of the ECSC Treaty). The resolution of 13 March was published in the Official Journal of 6 April (Official Journal 1981, C 77, p. 70) and the application was not lodged at the Registry until 7 August. Accordingly, even if, in accordance with the wellknown maxim magis valent quam pcreat, the application were taken to refer to the first of the two resolutions and it were held that it alone is in the nature of a measure which adversely affects the persons concerned and may be challenged, the application would be out of time because more than four months passed between the publication of the measure and the commencement of proceedings.

It is impossible to concur with this view. It is indeed true that, according to previous decisions of the Court, confirmatory measures may not be independently challenged (see for example the judgment of 22 March 1961 in Joined Cases 42 and 49/59 S NUPAT v High Authority [1961] ECR 53); however, it is also true that the resolution of 7 July does not merely confirm that of 13 March. On 13 March the Parliament, acting on a proposal from the Bureau, determined the calendar and places for the sessions for the whole of 1981, and more particularly, for the period from 23 Marcii to 18 December, deciding that in this period nine part-sessions should be held, all of them in Strasbourg (23 to 26 March, 6 to 10 April, 4 to 8 May, 15 to 19 June, 6 to 10 July, 14 to 18 September, 12 to 16 October, 16 to 20 November and 14 to 18 December). It is clear that that resolution is of limited scope because it is exclusively concerned with organizing the work of the full Parliament for that particular year. It contains no statement of a general nature on the procedures regarding the work of the Parliament in subsequent years; it does not consider, either in general or with reference only to 1981, the meeting places of the committees and political groups; finally, it does not adopt any measure concerning the operations of the Secretariat and of the technical departments.

The scope of the subsequent resolution of 7 July is entirely different. As the Court is aware, the Parliament determines in that measure that the plenary sittings will be held in Strasbourg, without stating how long that is to continue and without reference to the calendar for meetings for a specified period. Furthermore, it decides two matters which were not dealt with in the previous measure: it directs the meetings of committees and political groups to Brussels as a general rule and enacts measures concerning the operations of the Secretariat and of the technical departments. In short, one can state with confidence that the contents of the two resolutions are not even partially identical. Accordingly, the bar to the proceedings which was canvassed on that basis does not exist.

I would add that that result is supported by the case-law of the Court. In its judgment of 31 March 1971 in Case 22/70 (already cited) the Court held that a measure of the Council could not be regarded “as simply a confirmation of previous discussions” (in which case the proceedings brought against it would have been out of time) since a regulation adopted in the meantime by the Council had brought about a decisive change in the relevant field (cf. paragraph 66 of the decision). The point at issue in that case was the allocation of powers between the Communities and the Member States concerning negotiations on transport.

8. 

However, the Parliament challenges the admissibility of the application on yet another ground: the measure in question affects only its own administration and therefore may not be challenged.

The principle relied upon is properly stated. Measures of an institution which produce legal effects solely within that institution are immune from challenge. The Court has delivered a number of rulings to that effect. I will cite the judgment of 12 February 1960 in Joined Cases 16, 17 and 18/59 (Geitling and Others v High Authority [1960] ECR 17, at p. 25). In that judgment it was held that an action did not lie against the contested measure because, since it was not binding on the persons to whom it was addressed and did not bind the High Authority in the future performance of its tasks, the applicants could not be adversely affected by it. It was also held that the measure in question “contains no provisions giving rise to legal effects... it is simply an internal measure”. Accordingly, if the measure in question fell within the category of purely internal measures the objection of inadmissibility would have to be upheld.

However, there is reason to doubt whether the measure does fall within that category. If one looks at its objectives the contested measure is to be included amongst the so-called self-regulating measures whereby every institution governs its own internal structure and operations. For the moment I shall not raise the question whether that measure remains strictly within the framework of the self-regulating powers of the defendant institution or whether it exceeds them, thereby being vitiated by lack of competence. I shall restrict my consideration to a different aspect. As a rule measures of internal organization produce internal effects only, that is to say, effects limited to the administrative sphere of the body which issued them. However, there are exceptions to the rule. It is thus possible, and in fact it is not unusual, that such measures also produce external effects, thereby entailing a risk that the interests of other persons may be affected. An example of measures having both internal and external effect is furnished by the preliminary ruling delivered by the Court of Justice on 15 September 1981 in Case 208/80 (Lord Bruce of Donington v Aspden [1981] ECR 2205). In that case the question was raised whether the national authorities were bound by the decision taken by the Parliament, within the framework of measures of internal organization, to refund travel and subsistence expenses to its members on a lump-sum basis. The Court decided in the affirmative, refusing the Member States power to tax any income enjoyed by Members of the European Parliament by reason of their membership (see paragraphs 17 to 22). If that principle is applied to the present case it must be held that, although the objectives of the measure in question are concerned with internal organization, the measure might produce external effects capable of adversely affecting other persons and therefore Member States also. That suffices, in my opinion, to render the objection in question unfounded.

In view of this I do not consider it necessary at this stage to inquire further into the actual nature of the contested measure in order to establish whether it exhausts its effects within the framework of the institution or whether it also produces effects externally. That question is in fact closely connected with the questions raised by the submission that the measure in question is vitiated by lack of competence. They are two faces of the same coin. However, those matters are concerned principally with the substance of the case and I shall postpone my examination thereof until consideration of that latter aspect.

9. 

However, the principal argument whereby the defendant institution maintains that the application is inadmissible is different: that argument is based on the nature of the contested measure. Its basis is beyond question: the measure of 7 July 1981, it is stated, concerns the organization of the work of the Parliament and is therefore concerned exclusively with the activities of that institution, without distinguishing between the various powers conferred upon it by each of the three Treaties. It adds that such a measure cannot correspond to the words “act of the Assembly”, contained in Article 38 of the ECSC Treaty, and cannot therefore be contested. The argument is based, on the. one hand, on the finding that neither the EEC Treaty nor the EAEC Treaty envisages, expressly at any rate, that measures of the Parliament may be challenged and, on the other hand, on the view that Article 38 of the ECSC Treaty only permits measures whose subject-matter comes within its scope to be challenged. On that basis it is clearly necessary, according to the defendant, to hold immune from challenge measures concerning matters common to the three Treaties, as, for example, those concerning the organization and work of the institutions.

Of one fact there can be no doubt. The interpretation thereby placed on Article 38 is extremely narrow and if it is accepted very few of the measures enacted by the Parliament in its capacity as an ECSC institution may be challenged. If measures having no binding effect are excluded (which by itself considerably narrows the field since this Parliament is an institution whose task is principally supervisory: cf. Article 20 of the ECSC Treaty) three categories of measures remain:

Is the Parliament, then, practically immune from supervision? No. There are a number of reasons why the interpretation suggested by the defendant cannot be upheld. The first, of a general character, concerns the nature of the resolution. As the Court is aware, the Parliament adopted the measure in order to define the way in which its own work is to be carried out with regard to its meeting places and to the staff and technical assistance necessary. However, it ís surely impossible to break down the contents of the measure, distinguishing between the powers conferred upon the institution by each of the three Treaties and then, having made that dubious distinction, hold that the measure may be challenged under Article 38 of the ECSC Treaty only in so far as it relates to the powers conferred by that Treaty. Although such an operation may perhaps be contemplated in the abstract (but even here doubts may be entertained), it would be absurd in practice, since the Parliament is a single body and clearly can only function in accordance with uniform procedures, independently of the framework of powers laid down in the ECSC, EEC and EAEC Treaties.

The foregoing furthermore is confirmed by the type of internal organization which the defendant institution has enacted. Power to organize its own business is conferred upon the Parliament by separate provisions of the three Treaties; Article 25 of the ECSC Treaty, Article 142 of the EEC Treaty and Article 112 of the EAEC Treaty, all of which have the same content. On the basis of those provisions the Parliament has adopted uniform rules of procedure which apply to its operations relating to the ECSC (only Article 31 refers exclusively to powers based on that Treaty) and to those relating to the EEC and EAEC.

The Parliament does not deny that this is the position. In fact it recognizes these facts, but draws from them an argument to the effect that measures which concern the working of the institution, and as such are common to the ECSC, EEC and EAEC, may not be challenged under Article 38. Article 38, accordingly, must receive a narrow interpretation under which it would not apply to measures of the Parliament concerning also, but not exclusively, powers under the ECSC Treaty. In my view, however, the reasoning must be inverted: it is precisely the indivisible nature of the measure which prompts the view that it may be challenged. I arrive at this conclusion for various reasons.

It is necessary first of all to define the scope of Article 38 of the ECSC Treaty, independently of the subsequent EEC and EAEC Treaties, and of the Convention of 25 March 1957 on certain Institutions common tö the European Communities. I refer of course to the part in which the first Treaty provides that: “The Court may, on application by a Member State or the High Authority, declare an act of an Assembly... to be void.” It cannot be seriously doubted that these words also refer to measures whereby the institution regulates its own working (naturally in so far as such acts are capable of producing external effects). I am inclined to favour that view because of the complete clarity of the wording and, having regard to the interests protected by the provision, because of the irrational nature of the system which would result if we excluded from the categoiy of measures which may be challenged those concerning the working of the Parliament. There are no serious factors which militate in favour of such an exclusion or which render it plausible. On the contrary, the requirement that the law should be upheld, laid down in Article 31 of the ECSC Treaty (“the Court shall ensure that in the interpretation and application of this Treaty, and of rules laid down for the implementation thereof, the law is observed”) and confirmed by identical wording in the EEC Treaty (Article 164) and the EAEC Treaty (Article 136), indicates that the wide interpretation should be adopted.

If the foregoing is correct, in order to find for the defendant institution, it would be necessary to argue, as in fact it does, that the scope of Article 38 of the ECSC Treaty was reduced by the conclusion of the EEC and EAEC Treaties. I will remind the Court that the latter do not provide, at any rate expressly, that measures adopted by the Parliament may be challenged. Such a lacuna, it is argued, leads to the view that, in addition to measures relating exclusively to powers based on the EEC and EAEC Treaties, measures which concern common powers, including those affecting the structure and working of the institution, are not open to challenge. But that proposition does not withstand critical examination. There are provisions containing clear evidence to the contrary.

I refer first of all to Article 232 (1) of the EEC Treaty, which provides that: “The provisions of this Treaty shall not affect the provisions of the Treaty establishing the European Coal and Steel Community, in particular as regards the rights and obligations of Member States, the powers of the institutions of that Community and the rules laid down by that Treaty for the functioning of the common market in coal and steel.” It is clear that, if immunity from challenge is granted for measures concerning the working of the institution, two classes of powers are reduced: on the one hand, those of the Member States and of the High Authority, which are deprived of a form of judicial protection, and, on the other hand, those of the Court of Justice, which loses its power to review the lawfulness of measures which are often important. Such consequences are clearly at variance with the provision in question, the purpose of which is precisely to prevent the advent of the two new Treaties and their incorporation into a single framework from altering the legal position established by the ECSC Treaty for the institutions and the Member States. If the problems created by the introduction of new provisions and the need to coordinate them with the earlier provisions are appraised in accordance with the terms of Article 232 of the EEC Treaty, it must therefore be accepted that the EEC and EAEC Treaties have not modified Article 38, at any rate so far as the definition of measures which may be contested is concerned.

Article 30 of the Treaty of 8 April 1965 establishing a Single Council and Single Commission of the European Communities proceeds along the same lines. It states: “The provisions of the Treaties establishing the European Economic Community and the European Atomic Energy Community relating to the jurisdiction of the Court of Justice and to the exercise of that jurisdiction shall be applicable to the provisions of this Treaty..., with the exception of those which represent amendments to articles of the Treaty establishing the European Coal and Steel Community, in respect of which the provisions of the Treaty establishing the European Coal and Steel Community shall remain applicable.” The provision is plainly prompted by the requirement which, as we have just seen, is intended to comply with Article 232 of the EEC Treaty, namely to leave unaltered the powers conferred upon the Court of Justice by the ECSC Treaty, preventing the merger of the executive bodies from affecting them, just as the establishment of the new Communities and the contemporaneous unification of the Assemblies and of the Courts did not affect them.

The view may accordingly be taken that the provisions subsequently enacted did not amend Article 38 of the ECSC Treaty. The same view was, moreover, expressed by Mr Advocate General Reischl in his Opinion delivered in Case 66/76 (CFDT v Council [1977] ECR 311, at p. 314). It is thus confirmed that the provision in question still provides that an action lies against measures of the Parliament, including those concerning the working of that institution.

In support of its reading of Article 38 the defendant institution then advances an argument based on what is alleged to be the actual intention of the contracting parties, which comes close to suggesting that the latter should take precedence over the provisions of the Treaties. It is observed (in the rejoinder dated 19 February 1982, at p. 13) that the EEC and EAEC Treaties — which do not envisage, at any rate expressly, that measures of the Parliament may be challenged — were adopted after the ECSC Treaty and accordingly set out the most recent point of view of the Member States on the question of review by the Court of the actions of the Parliament. It is argued that, having regard to that indication, it would be reasonable to place a restrictive interpretation on Article 38. However, that argument too must be rejected — it encounters an insuperable obstacle in the clear wording of Article 232 of the EEC Treaty. I consider that for the purpose of interpretation recourse may be had to a subjective factor such as that relied upon by the Parliament only where provisions are ambiguous. In this case there is no shadow of ambiguity.

In any event, the interpretation of Article 38 which I propose finds support in the case-law of the Court. As we have seen, in its judgment of 15 September 1981 in the case of Lord Bruce of Donington v Aspden (cited above), the Court considered the provisions laid down by the Parliament governing the reimbursement of costs and allowances for its own Members. Although the operative part of the judgment makes a general reference to “Community” law, the grounds of the judgment contain an examination of the organizational measure governing payment of travel and subsistence expenses (paragraph 15). Since it concerns the working of the institution that measure was clearly adopted in exercise of powers conferred upon the Parliament by all three Treaties; that is to say, it was a typical example of a “common” measure. The ECSC Treaty confers upon the Court jurisdiction to give preliminary rulings on the validity (but not on the interpretation) of measures of the High Authority and of the Council alone (Article 41 of the ECSC Treaty). On the other hand, the EEC Treaty confers upon it jurisdiction to give preliminary rulings “concerning... the validity and interpretation of acts of the institutions of the Community” (Article 177 (b)), thus including measures adopted by the Parliament. When measures of the Parliament are involved which, relating to the working of the institution, are “common” to the Treaties, the same problem of construction as that created by Article 38 of the ECSC Treaty arises in reverse terms.

In these cases the Court is faced with a very abrupt choice: either it must accord Article 177 a wide scope embracing common measures as defined above or it must entirely exclude such measures from the supervision which it exercises by way of preliminary rulings. As early as Case 101/63 (Wagner v Fobrmann and Krier [1964] ECR 195) the Court was required to consider a similar problem: whether it had jurisdiction to deliver preliminary rulings on questions referred to it under Article 177 of the EEC Treaty in so far as such questions concerned the interpretation of the ECSC Treaty. The Court decided the case without settling that point; however, the Advocate General expressed a clear view. In its judgment in the Lord Bruce of Donington case, on the other hand, the Court opted, albeit in terms which are by no means explicit, for a wide interpretation of Article 177.

This tendency is developed from a statement of a general nature made in the judgment of 31 March 1971 in Case 22/70 (Commission v Council, already cited). The Court was required to decide whether under Article 173 of the EEC Treaty the Member States and the institutions may only challenge the measures having binding effect which a.re specified in Article 189 of the EEC Treaty or “all measures adopted by the institutions, whatever their nature or form, which are intended to have legal effects” (cf. paragraphs 38 to 43 of the decision); the Court opted for the latter alternative, which requires a wide interpretation of Article 173 of the EEC Treaty.

It is a choice, which in my view, must be upheld. At its basis — and it is worth emphasizing this — lies the need to ensure, in accordance with the provisions of the Treaties, the widest and most thorough supervision by the Court of the lawfulness of the measures adopted by the Community institutions. In the case of the Parliament that need is more obvious today than it was in the past because the Parliament, by reason of its new representative status, carries greater weight, which will presumably increase in the future. Andrew Jackson wrote, in a letter to Benjamin Franklin, that the guarantee of American freedom resided in the jealousy between the courts, the President and Congress. Jealousy in this case is not the ignoble sentiment which is disapproved of in matters of general morality: it is the motive force of the system of checks and balances whereby the pluralist democracies govern themselves with their catalogue of fundamental rights. I venture to suggest that the Court should encourage these views.

10. 

Article 38 of the ECSC Treaty therefore gives the Court of Justice jurisdiction to entertain this application. The article is both necessary and sufficient for jurisdiction. It is thus unnecessary to deal with the argument which endeavours to provide additional ground for jurisdiction on the basis of Article 173 of the EEC Treaty by interpreting it widely. Furthermore, the applicant put forward that argument in the alternative.

11. 

I shall now proceed to consider the substance of the case. The Grand Duchy claims that the resolution of 7 July 1981“on the seat of the institutions of the European Community and in particular of the European Parliament” is vitiated by lack of competence and infringement of essential procedural requirements. On the one hand, it is stated, the three Treaties provide that the decision on the seat of the institutions is to be taken by the governments of the Member States; on the other hand, it is added, in the procedure for approving the measure the Legal Affairs Committee was not consulted. The Court is empowered to take cognizance of both of the grounds of invalidity under the last paragraph of Article 38 of the ECSC Treaty, which, with reference to an application for a declaration that an act of the Assembly is void, provides that “the only grounds for such application shall be lack of competence or infringement of an essential procedural requirement”.

I shall deal first with the alleged infringement of an essential procedural requirement. In my opinion, the failure to consult the Legal Affairs Committee does not affect the validity of the contested resolution. The Parliament, in regulating its own proceedings, as is authorized by Article 25 of the ECSC Treaty, Article 142 of the EEC Treaty and Article 112 of the EAEC Treaty, has assigned to the committees the task of examining questions submitted to them by the Parliament. It is true that the rules refer to the matters to be dealt with by -the various committees and envisage conflicts of jurisdiction between those committees; it appears to me, however, that the question whether those rules are complied with can have no external relevance. In this matter the final word must surely lie with the defendant institution, whose choices, guided as they are by appraisals which are largely a matter of discretion, clearly cannot be subject to the review of the Court. Furthermore, in this case the Political Affairs Committee, which was considered the most appropriate body for considering the location and organization of the work of the Parliament, was heard: there were good grounds for that if it is accepted that the question of the seat is primarily, if not wholly, of a political nature.

12. 

I now come to the issue of lack of competence. The applicant claims that when the Parliament adopted the resolution of 7 July 1981 it substituted itself for the governments of the Member States in the exercise of the powers reserved to them in choosing the seat of the institutions.

In order to ascertain whether that complaint is well founded, and if so to what extent, it is first of all necessary to rehearse the various agreements from 1952 to the present day which have been concluded between the governments with regard to the location of the institutions and their departments. Then the terms of the resolution of 7 July 1981 must be analysed and the question must be put whether or not the resolution is in accordance with the agreements.

Two points, however, must be clarified now. It has been stated that the three Treaties reserve to the Member States power to decide on the seat of the institutions. The relevant provisions are Article 77 of the ECSC Treaty, Article 216 of the EEC Treaty and Article 189 of the EAEC Treaty, which I have already considered above. All of these articles refer to the “seat of the institutions” without distinguishing between provisional seats and definitive seats. The question may accordingly be raised whether the Member States have exclusive powers also in determining the provisional seats. I consider that they undoubtedly do. In the first place, the very wording which I have quoted points to that conclusion: if the various provisions employ the word “seat” and do not qualify it, it is highly probable that they were intended to cover both cases. The probability becomes a certainty when to the argument based on the wording is added an argument a contrario. If the words were interpreted narrowly the reservation of power to the Member States would be weakened: by means of decisions represented as being temporary and thus nominally subordinate to a future, definitive agreement between the governments, the institutions would in fact ultimately deprive the governments of a power which is conferred upon them by the Treaties. That is plainly at variance with the clear objective of the provisions: to deprive the Community institutions of any power in the choice of the seats.

The abovementioned provisions then give rise to a second query; what must be understood by the “seat” of an institution? Certain sources refer to “provisional places of work”; that is the case, for example, in Article 1 of the “Decision of the representatives of the Governments of the Member States on the provisional location of certain institutions and departments of the Communities”, which provides that: “Luxembourg, Brussels and Strasbourg shall remain the provisional places of work of the institutions of the Communities” (my italics). The Parliament's resolution, on the other hand, repeatedly emphasizes the need for a “single working place” (cf. in particular paragraph 2) and puts forward practical reasons therefor. In my view, the seat and the place of work are completely identical; there is no Community language in which “seat” (sede, siege, Sitz and zetel) does not mean the place where an organization has its own offices and usually carries on its business. It is probable that the said decision of the representatives of the governments refers to “places of work” because the choice made in that decision was expressly provisional and in a situation of that nature the word “place” appeared less binding than the word “seat”. Furthermore, in the preamble to the same decision it is stated that that choice was made “without prejudice to the application of Article 77 of the Treaty establishing the European Coal and Steel Community, Article 216 of the Treaty establishing the European Economic Community, Article 189 of the Treaty establishing the European Atomic Energy Community...”. How can one escape the conclusion that the decision was concerned with the seat, although it did not settle the basic problem, thereby preserving intact the powers of the governments.

13. 

Now that I have dealt with these problems of definition I shall endeavour to describe the terms of the agreements concluded by the governments concerning the seat of the Parliament and the practice which has developed in this matter.

At a press conference held on 24 and 25 July 1952, to coincide with the entry into force of the ECSC Treaty, the Foreign Ministers of the six Member States declared that the High Authority and the Court of Justice would begin their work in Luxembourg and that the Assembly would hold its first, session in Strasbourg. They then added that the final choice of a seat would be made in the light of the negotiations to be undertaken concerning the status of the Saar (cf. Annex 1 to the Parliament's defence dated 14 October 1981). Following that decision the sessions of the Common Assembly of the ECSC not only began but continued, with very few exceptions, to be held in Strasbourg; for its part, the General Secretariat began to operate in Luxembourg because the Council met there and the High Authority and its departments were established there. That situation continued unchanged until the entry into force of the Treaties of Rome.

The second stage commenced on 2 January 1958. At a meeting in Paris the Foreign Ministers of the Member States agreed to bring together in the same place all of the European organizations of the six countries as soon as that concentration became practical and in accordance with the provisions of the Treaties. In order to select the seat they decided to meet again before 1 June 1958. On the same occasion they agreed that the meetings of the Commissions should be convened by their Presidents. Furthermore, having regard to the fact that no seat, final or provisional, had been decided upon the governments recommended that the Commissions should hold their meetings at Val Duchesse (Brussels) or in Luxembourg for practical reasons and because of the available facilities. Finally, the Ministers decided that the Assembly should sit in Strasbourg (cf. press release concerning the conference in Paris of the Foreign Ministers dated 7 January 1958, Annex 2 to the Parliament's defence dated 14 October 1981).

The press release which I have just quoted bears the significant title of “seat” and shows that the governments of six countries had arrived at an agreement in considerable detail on the place where the Assembly was to work. The agreement provided expressly that the Assembly should meet in Strasbourg. On that point, then, the indication given by the Ministers for Foreign Affairs at the conference of 24 and 25 July 1952 was consolidated: at that time it had merely been said that the Assembly /would hold its “first session” in Strasbourg; in the later document it was stated that the sittings would be held in Strasbourg, without a limit as to time. The new agreement in fact reflected the practice which had developed in the previous years on the basis of the cautious (or merely allusive) statement of 1952. I will add that at the same ministerial conference in 1958 the work of the Commissions was also considered; however, it was only “recommended” that they should meet in Brussels or in Luxembourg.

In accordance with all these agreements the Assembly continued to hold its sessions in Strasbourg, whilst the Secretariat continued to function in Luxembourg. With regard to the meetings of the parliamentary committees, the settled practice developed of holding them principally in Brussels where the executive bodies of the European Economic Community and of Euratom were established.

The question must be raised at this point as to the status of the agreements between the governments which I have referred to. The reply proceeds on the assumption that, with regard to the seats of the institutions, and on the basis of an express reservation of powers laid down in the Treaties, the Member States are acting outside the Community framework and are accordingly governed by the traditional rules on the conclusion of agreements. On that basis it does not appear to me that the agreements may be regarded as treaties in simplified form and accordingly capable of entering into force without the need for ratification through the signature of plenipotentiaries. In my opinion, the agreements are not of a legal nature. I am led to this view principally by the circumstance that their contents are given only in press releases; the adoption of such a form in fact casts serious doubt on the intention of the parties to create specific legal obligations.

14. 

The location of the Parliament — this is the third stage — was subsequently defined seven years later within the framework of the negotiations for the establishment of a single Council and a single Commission of the European Communities. The agreement concluded at the ending of the negotiations (on 8 April 1965) provides in Article 37 that, “without prejudice to the application” of Article 77 of the ECSC Treaty, Article 216 of the EEC Treaty, Article 189 of the EAEC Treaty and Article 1 of the Protocol on the Statute of the European Investment Bank (the second paragraph of which states that: “The seat of the Bank shall be determined by common accord of the Governments of the Member States”), “the representatives of the Governments of the Member States shall by common accord lay down the provisions required in order to settle certain problems peculiar to the Grand Duchy of Luxembourg which arise out of the creation of a single Council and a single Commission of the European Communities”. The second and last paragraph of the same article adds: “The decision of the representatives of the Governments of the Member States shall enter into force on the same date as this Treaty.”

On the basis of the provision cited (but not, as will be seen, on that basis alone) the representatives of the Governments of the Member States adopted on the same day as the Merger Treaty a decision comprising 13 articles “on the provisional location of certain institutions and departments of the Communities”. As was stated in the preamble, the contracting parties, in taking that decision, recognized that at the time of the merger of the executive bodies and “in order to settle certain problems peculiar to the Grand Duchy of Luxembourg” (I note in passing that this passage reproduces almost to the letter the wording of Article 37 of the basic agreement) it was appropriate “to designate Luxembourg as the provisional place of work of certain institutions and departments”.

The principal provisions of the decision relating in some degree to the activity of the Parliament are Articles 1, 4 and 12. Of those provisions the most relevant to our purposes is Article 4. It provides : “The General Secretariat of the Assembly and its departments shall remain in Luxembourg.” I must first of all point out that a similar provision is not contained in the agreements of 1952 and 1958. From 1958, however, it was precisely that practice which developed, whicli explains the use of the words “shall remain”. On the other hand, Article 1 is of general scope and refers to the provisional places of work of all the Community institutions, including the Parliament; it provides: “Luxembourg, Brussels and Strasbourg shall remain the provisional places of work of the institutions of the Communities.” The provision must be read in context, that is to say with reference to the previous agreements between the governments, which, so far as they are relevant to this case, designated Strasbourg as the location for plenary sessions. That the decision must be interpreted in the light of the agreements emerges even more clearly from Article 12, which provides: “Subject to the preceding provisions, this decision shall not affect the provisional places of work of the institutions and departments of the European Communities, as determined by previous decisions of the Governments, nor the regrouping of departments occasioned by the establishing of a single Council or a single Commission.”

The conclusion may thus be drawn that the agreements of 1958 have acquired a binding nature which they did not initially possess, through the reference made to them by the decision of 1965.

15. 

However, the Parliament endeavours to confer a narrow interpretation on that decision, emphasizing its connection with Article 37 of the Merger Treaty. Such a connection, it is agreed, would exclude the possibility that the Member States adopted the decision in exercise of the power to choose the seat of the institutions as laid down in the Treaties of Paris and of Rome: it is clear that a decision divorced from the Treaties would not be binding on the defendant institution (cf. p. 28 of the rejoinder, lodged on 23 February 1982). But there are many considerations which preclude the adoption of the argument thus summarized.

In the first place, it is extremely questionable whether the terms of the decision are at variance with Article 37. As the Court is aware, the latter provision requires the representatives of the governments to lay down “the provisions required in order to settle certain problems peculiar to the Grand Duchy of Luxembourg...”. In fact all of the articles of the decision (except Article 13 which is of a procedural nature and concerns the time of the entry into force of the decision) relate, wholly or in part, to specific Luxembourg interests. It is true that some provisions take into consideration the interests of other countries: for example Article 1, which mentions Brussels and Strasbourg (besides Luxembourg) as provisional places of work of the institutions of the Community and Article 6, which gives Brussels (again, besides Luxembourg) as the place where the Monetary Committee must meet. It should be noted however, that these provisions too merely clarify the position of Luxembourg, evidently with a watchful eye on the conflicting interests.

Let us none the less accept that the decision exceeds the limits laid down by Article 37; that does not provide grounds for considering it invalid for failure to comply with the assumed delegation of power and, accordingly, incapable of binding the Parliament. In fact, in my view, its legal basis lies not so much, and not exclusively, in Article 37 as in Article 77 of the ECSC Treaty, Article 216 of the EEC Treaty and, so far as it is concerned with the seat of the European Investment Bank, to which Article 5 refers, Article 1 of the Protocol on the Statute of the European Investment Bank. It thus appears to me impossible to consider that there has been a proper delegation of powers as such and then to assume that it has not been complied with. The Parliament does not concur. It observes that Article 37 opens with the words “without prejudice to the application of” Article 77 of the ECSC Treaty, Article 216 of the EEC Treaty, Article 189 of the EAEC Treaty and Article 1 of the Protocol on the Statute of the European Investment Bank; and that reservation is such as to exclude the possibility that the governments, in adopting the decision in question, were acting in exercise of the powers conferred upon them by the primacy sources.

But that is not the case. I for my part consider that the reservation (subsequently repeated in the preamble to the decision) is to be explained much more simply and logically: when the States concerned introduced the reservation they merely intended to make clear that the choices made were provisional and that they retained their right to determine the seats of the institutions definitively and perhaps differently. It is accordingly impossible- to employ the reservation as a means of denying the relationship between the decision of 1965 and the powers of the governments based directly on the Treaties of Paris and of Rome. Nor is it possible to urge against that relationship that the decision was intended solely to determine (albeit partly) the provisional places of work of the institutions. I have already said that the provisions of the Treaties do not merely deal with the definitive seats, the provisional places of work are also covered.

Finally, I do not consider that the form of the decision may be relied upon in order to justify the argument of the Parliament. Why the form? Because whilst the Merger Treaty entered into force through ratification (cf. Article 38) the decision attained the force of law by other means. The second paragraph of Article 37 of the Merger Treaty in fact provides: “The decision of the representatives of the Governments of the Member States shall enter into force on the same date as this Treaty.” An identical requirement is contained in Article 13 of the decision itself. That connection between the decision and the Treaty strengthens the impression that the former is subordinate to the latter and again tends to exclude the possibility that the governments, in adopting the former, were acting in exercise of a power directly based on the ECSC, EEC and EAEC Treaties.

The argument is, however, fragile. All that may be inferred from the procedure laid down for the entry into force of the decision is that it is in the nature of an agreement in simplified form, that is to say that legal effect is conferred upon it by the signature of the plenipotentiaries without the need for ratification. The connection which the provisions cited above establish between the two sources is thus of reduced importance: it merely implies that the procedure for concluding the agreement was completed by signature and that its entry into force was dependent, for obvious reasons concerning coordination, on the entry into force of the Merger Treaty.

16. 

I shall now consider in what places the Parliament has functioned in the period extending from April 1965, the date of the decision with which I have just dealt, and the commencement of the fourth stage of the events in question, in March 1981, when the Heads of State and Heads of Government of the Member States, meeting in Maastricht as the European Council, set out their views on the seat of the institutions.

For a further two years the sessions were held in Strasbourg. It was not until 19 July 1967 that for the first time that custom was interrupted: on the basis of a decision taken by the Bureau (Document PE 17.995, BUR, Annex 4 to the Parliament's defence of 14 October 1981) the Parliament held a sitting lasting one day in Luxembourg. Subsequently short sittings were generally held there. A table produced by the defendant and not disputed (Annex 5 to the defence of 14 October 1981) shows that in the period from 1966 to 1978 the Parliament held 87 sittings in Strasbourg amounting to a total of 409 days and 57 sittings in Luxembourg amounting to a total of 190 days. At a certain point France reacted: by a letter of 4 February 1971 addressed to the President of the Parliament (Annex 6 to the defence) the French Minister for Foreign Affairs expressed the concern of his Government. In 1971, as a result of a decision of the Bureau, five out of 11 sessions of the Parliament were held away from Strasbourg so that, as the Minister observed, what were previously exceptional cases were becoming regular. The Minister also stated that such a practice was not in accordance “either with the Treaties of Paris and of Rome, which reserve to the Governments of the Member States power to determine the seat of the Communities, or with the decisions which those Governments have taken to determine the provisional places of work of the Community institutions”. In fact, the Minister continued, “those decisions — that of 7 January 1958, in accordance with which the Assembly must meet in Strasbourg, and that of April 1965, which confirmed the previous decision on that point — do not leave any doubt that Strasbourg is the provisional place of work of the Assembly, at least so far as the sessions are concerned”.

On 8 March 1971 (see Annex 7 to the defence) the President replied to the French Government explaining that the decision to hold certain sessions in Luxembourg as well as Strasbourg was dictated by two practical difficulties: that of transferring from time to time departments of the Secretariat from Luxembourg to Strasbourg for the duration of the sessions and that of reconciling the work of the Commission, whose “usual place of work is Brussels”, with the functioning of the Parliament in Strasbourg.

The French Minister for Foreign Affairs again wrote to the President of the Parliament on 26 January 1973 (Annex 8 to the defence) expressing his Government's concern over the Parliament's practice “of holding each year a certain number of its sessions in Luxembourg”; the more so because according to the calendar for 1973 for the first time a number of such sessions were of long duration. He emphasized that such a practice was not in accordance either with the Treaties of Paris and of Rome or with the decisions adopted by the Governments in 1958 and 1965.

In 1977 the President of the Council of Ministers of the Communities set out his views on the matter. On 22 September the President of the Council, in reply to a letter sent to him by the President of the Parliament (Annex 9 to the defence) pointing out the practical requirements arising from the functioning of the Parliament in three separate places — Strasbourg, Luxembourg and Brussels — declared that the governments of the Members States did not consider that there was “reason to modify either in law or in fact the provisions at present in force concerning the provisional places of work of the Assembly... These places are Strasbourg and Luxembourg, where its General Secretariat and its departments remain established”. In the same letter the President stated: “It is, furthermore, common knowledge that the parliamentary committees are accustomed to meet in Brussels, with the minimum level of staffing required for holding their meetings” (Annex 10 to the defence). The views thus set out by the Council met with the approval of the Luxembourg Government, whose President stated inter alia in a letter of 19 January 1978 (Annex 12 to the defence): “On the basis of the decision of the representatives of the Member States on the provisional location of certain institutions, dated 8 April 1965, the provisional places of work are Strasbourg and Luxembourg, where the Secretariat of the Assembly and certain departments have been established.”

On 22 September 1978 a new protest followed from the French Government. In a letter addressed to the President of the Parliament the Minister for Foreign Affairs again expressed grave concern over the decision taken on 14 September 1978 by the Bureau of the Parliament, which, in drawing up the calendar for the first six months of 1979, provided that a certain number of sessions should be held in Luxembourg. He repeated that according to the decisions of 1958 and of 1965 Strasbourg was “the sole meeting place for the Assembly” and continued that “the practice which has gradually developed of holding certain short sessions in Luxembourg is contrary to those decisions and has not at any time been approved by the governments” (Annex 15 to the defence).

Reference must also be made to a letter of 10 October 1978 sent by the President of the Council of Ministers of the Communities to the President of the Parliament in which he stated that “the governments of the Member States did not consider that there was any reason to modify either in law or in fact the provisions at present in force concerning the provisional places of work of the Assembly” (Annex 16 to the defence).

Finally, in 1979 the Parliament was elected by direct universal suffrage. It held its first sessions between July of that year and June 1980 in Strasbourg; between June 1980 and February 1981 it also held four sessions in Luxembourg, where in the meantime a large new chamber had been constructed. The direct election and the extraordinary increase in the number of Members of Parliament rendered more serious and urgent the problem of the place of work, so much só that, in a memorandum of 16 September 1980, the French Government proposed to the “Governments of the other Member States that they should begin without delay talks in order to resolve this problem, pursuant to Article 77 of the ECSC Treaty, Article 216 of the EEC Treaty and Article 189 of the EAEC Treaty (Annex 5 to the application).

Partly as a result of that initiative, the defendant institution adopted on 20 November 1980 a resolution “on the seat of the European Parliament” in which it called upon the governments of the Member States to decide the question by 15 June 1981 and issued the following warning: “If the governments of the Member States have not reached a decision by the above date it would have no option but to take the necessary steps to improve its working conditions” (Official Journal, C 327 of 15 December 1980, p. 49). Accordingly, on 12 January 1981 the Parliament rejected the draft calendar put forward by the Bureau, which provided, for the first half of 1981, that two part-sessions should be held in Luxembourg in accordance with the practice followed in recent years. It subsequently approved a different calendar, at the plenary session of 13 March. According to the hew calendar, in the period from 23 March to 18 December 1981 the part-sessions were to be held exclusively in Strasbourg (Official Journal, C 77 of 6 April 1981, p. 70).

That decision provoked a reaction from Luxembourg. On 17 March 1981 the Luxembourg Minister for Foreign Affairs addressed a letter to the President of the Parliament expressing his Government's concern over the fact that no part-session would be held in Luxembourg in the second half of 1981 (Annex 18 to the defence).

However, the step taken by the French Government in September 1980 also produced other effects. On 26 and 27 March 1981 the Heads of State and Heads of Government, meeting at Maastricht as the European Council, considered the problem of the seat of the institutions. The views which they adopted at that meeting were published in the Bulletin of the European Communities. The communiqué reads: “The Heads of State and of Government decided unanimously to confirm the status quo in regard to the provisional places of work of the European institutions” (Bulletin, 1981, No 3, p. 9).

Subsequently, at the end of June 1981, the inconvenience caused to the Parliament through having to work in different places prompted the representatives of the governments to hold a meeting. The conclusions reached at the meeting were as follows: “1. The Governments of the Member States find that, in accordance with Article 216 of the Treaty, they alone are empowered to determine the seat of the institutions of the Community. 2. The decision of the Governments of the Member States taken at their meeting at Maastricht on 23 and 24 March 1981 to maintain the status quo in regard to the provisional places of work comes within the exercise of that power. It is without prejudice to establishing the seat of the institutions...” (Annex 7 to the application).

Finally, on 7 July 1981, that is to say, only a few days after the meeting which I have mentioned, the Parliament adopted the resolution which forms the subject-matter of this case.

17. 

I shall shortly analyse the content and scope of that resolution. Let us first raise the point, having regard to the practice which has developed since 1965 and the agreements concluded at Maastricht and at the meeting at the end of June 1981, whether new facts have supervened which affect to some degree the provisional location of the Parliament as laid down in the decision of 1965 and in the preceding agreements (on this point, see in particular section 14, supra). I must state from the outset that in my view the legal position has not altered. The rights and duties of the Member States, on the one hand, and the position of the defendant institution, on the other, remain those clearly defined in the said decision.

It would be convenient to repeat that, in accordance with that decision and a practice which is now well established, the location of the departments and work of the Parliament is as follows: sessions are held in Strasbourg; Luxembourg is the seat of the Secretariat and of the departments; the Parliament holds committee meetings in Brussels. Within that framework the most significant innovation between 1965 and 1981 is the choice of Luxembourg, in addition to Strasbourg, as the place for holding sessions. Is it possible for that practice and the subsequent agreements of March and June 1981 to be considered sources in the legal sense and thus capable of requiring some of the plenary sessions to be held in Luxembourg?

As I have pointed out, I do not think that that is the case, for a number of reasons. For example, the practice of holding certain sessions in Luxembourg is, I consider, subject to too many restrictions for it to affect a legal position founded, as the previous one was, on a formal decision. Opinions differ with regard to the principles governing the relationship between written provisions and custom. Some authorities doubt whether the terms of a multilateral treaty (and the agreement of 1965 may be classified as such) may be modified purely as a result of a course of conduct unanimously adopted by the contracting parties over a period of time. However, in the case of the practice in question not even these conditions obtain. It is sufficient to consider that the course of conduct has lasted for only 13 years (from 1967 to 1980) and not all the States concerned have adhered to it. It is significant that at least one State, France, has directed many protests to the President of the Parliament insisting that the Paris agreements of 1958 and the decision of 1965 prescribe Strasbourg alone as the provisional place for holding sessions. One thinks of the letters of the French Minister for Foreign Affairs to the President of the Parliament dated 4 February 1971, 26 January 1973 and 22 September 1978, the memorandum of the French Government of 16 September 1981 and the two corresponding letters from the President of the Council of Ministers (22 September 1977 and 10 October 1978).

Nor can the agreements of March and June 1981 be considered to bring about any alteration. In this respect the decisive factor appears to me the extremely succinct words (“confirm the status quo”) of the agreement concluded in Maastricht, whilst nothing was added to it by the meeting at the end of June, which merely stated that the Maastricht decision was adopted by the governments in exercise of the power reserved to them by Article 216 of the EEC Treaty. I have described these words as “succinct”; precisely for the same reason I could, however, have also called them ambiguous, or rather, intentionally ambiguous. What in fact did the governments mean when they referred to the “status quo”? The previous agreements or the practice which developed subsequently? I consider that in the absence of express or indirect evidence to the contrary the Maastricht agreement must be interpreted in accordance with the agreements of 1965. If the governments had intended to modify those agreements in any way they would have employed words which were less obscure and less open to various interpretations and would perhaps have cast their agreement in a different form. I am not unaware — I should even say I am certain — that some of the countries concerned were induced to accept the words “status quo” in the belief that they were approving an innovation of the old system. However, it is no less certain that many — the majority — of the countries agreed because they believed that the decision of 1965 would not be affected. In that state of affairs I consider that one, and only one, interpretation is reasonable: to regard the agreements of 1981 as confirming those of 1965.

18. 

It must now examine the resolution adopted by the Parliament on 7 July 1981 (Official Journal, C 234 of 14 September 1981, p. 22). In this connection, the first requirement is to review the various aspects of the resolution. Thereafter it will be necessary, in the light of the results of that review, to establish whether, or to what extent, the resolution is compatible with the agreements concluded by the Member States on the seat.

I shall briefly summarize the terms of the resolution. The preamble, which is very extensive (it consists of 23 recitals), rehearses the various events concerning the seat, from the decision of 1965 to the meeting in Maastricht, and points out the objective difficulties which the defendant institution encounters in its work through the lack of a single place of work. It emphasizes furthermore that under the Treaties the Member States have “the right and duty to determine ‘the seat of the institutions of the Community by common accord’”(the first recital of the preamble); it requests (like the previous resolution of 20 November 1980, which has already been cited) that the Parliament should be consulted before the decision on the seat is taken and observes that that procedure” — by analogy with the existing procedure for conciliation between the European Parliament and the Council — is justified by the extremely important financial consequences of the fixing of the seat” (the nineteenth recital).

The operative part of the resolution is drawn up in four paragraphs: the first “calls on the Governments of the Member States to comply with their obligation under the Treaties and... fix a single seat for the institutions of the Community” in accordance with a conciliation procedure; in the second paragraph the Parliament declares that “it is essential to concentrate its work in one place”; in the third paragraph it decides on a series of measures to be adopted provisionally “pending a final decision on a single meeting place of the European Parliament”; finally, in the fourth paragraph the President is instructed to forward the resolution to the governments of the Member States, the Council and Commission, and the other Community institutions. It is, of course, the third paragraph on which we must concentrate our attention. It is itself divided into three subparagraphs (denoted by the letters (a), (b) and (c)), which we must examine separately.

19. 

However, before we carry out that examination it is necessary to consider a question of a general nature. According to the applicant, in the resolution at issue the defendant institution has exceeded the power of organizing its own administration and its own work which it possesses, like all the other institutions. But in order to prove this the applicant invokes the context in which the resolution was adopted rather than its content. It urges that the minutes of the Parliament's proceedings should be consulted: they show clearly that the Parliament did not intend to abide by the agreements concluded by the governments, above all with regard to the redistribution of staff. I wish to state from the outset that this approach does not appear to me to be correct.

The minutes of the Parliament's proceedings do not in fact throw any light on the resolution. The examination carried out by the applicant leads to the conclusion that the minutes contain everything and nothing. For example, many Members of Parliament did indeed think that they were deciding the matter of the seat; and the warning contained in the abovementioned resolution of 20 November 1980 may indirectly create the impression that the Parliament has usurped the role of the governments in the choice of the seat, or rather seats. But there is no lack of evidence to the contrary. The Parliament, for example, rejected a series of amendments which, had they been adopted, would clearly have required the resolution to be classified as a decision on the seat. The sum total of so many ambiguities is nil. Furthermore, when a measure originates in an assembly it should be interpreted, so far as possible, in reliance on the text approved by the assembly.

I now proceed to examine the resolution in detail.

20. 

Paragraph 3 (a) sets out the Parliament's decision “to hold its part-sessions in Strasbourg”. These words imply that it is no longer intended to hold sessions in Luxembourg, at least until such time as an amendment to the measure in question is adopted. May it be said that that subparagraph constitutes a breach of the reservation of power to the Member States referred to in Article 77 of the ECSC Treaty, Article 216 of the EEC Treaty and Article 189 of the EAEC Treaty? I consider that it may not and I shall explain why.

I have already shown that, under the agreements hitherto concluded between the governments of the Member States, Strasbourg is the Parliament's provisional place of work as far as the sessions are concerned. It is therefore impossible to charge the Parliament with having usurped the power of the Member States in deciding to hold sessions in Strasbourg; by the decision contained in paragraph 3 (a) it has merely accepted and brought into force the decisions adopted by the governments. It is indeed true that that decision to some degree disappoints the expectations of the applicant State. The Court is aware that from 1967 a certain number of part-sessions have been held in Luxembourg each year; thus the change in policy which that paragraph of the resolution entails disregards Luxembourg's interests. However, those interests are not such that the Court must take cognizance of them. The agreements between the governments on the seat of the institutions, in particular that, or rather those, concerning the Parliament, do not in fact recognize Luxembourg as a place for holding sessions; nor, as has already been shown, is the practice which developed between 1967 and 1980, capable of modifying those agreements.

One point should, however, be made. What I have just said does not imply that the decision to hold certain part-sessions in places other than Strasbourg is contrary to the agreements concluded in 1965 and confirmed in 19.81. There is no breach of those agreements in so far asthe decision employs powers of self-regulation based either on the general principles governing the working of all public organizations or on the provisions of the Treaties empowering the Parliament to draw up its own rules of procedure (Article 25 of the EÇSC Treaty, Article 142 of the EEC Treaty and Article 112 of the EAEC Treaty). Disregarding the particular features of specific cases, we may say in short that a measure like the decision in question (and in future there may be more of them) is lawful if it fulfils two conditions: the first is that particular circumstances exist which justify the decision and which are based principally on the operative requirements of the institution; the second is that the number and duration of the part-sessions held outside Strasbourg should not attain such dimensions as to distort the meaning of the intergovernmental agreements in force.

21. 

Paragraph 3 (b) contains the decision “to organize the meetings of ... committees and political groups as a general rule in Brussels...”. Since I have accepted a wide definition of “seat” as the place (or places) where the institution has its own offices and usually works, I naturally consider that the location of such offices and work also falls within the powers of the governments. So in this respect also the resolution does not, in my view, depart from the agreements between the governments; I therefore consider that the submission of lack of competence pleaded by the Grand Duchy is unfounded.

I shall now endeavour to review the contents of the abovementioned agreements with regard to the place where the committees and political groups work. It has already been seen that the agreements of 1958 take into consideration only one of the Parliament's activities — the plenary sessions. However, between 1958 and 1965 (and subsequently until 1981) the committees usually met in Brussels. They did so because Brussels is the seat of the Community's executive bodies — with whose work the work of the committees is very closely connected. It is thus a sensible practice and, what is of greater significance, it is one which in the 23 years of its existence has never been challenged by the Member States. I deduce from this that what occurred is an implied agreement whose legal status does not differ from that of the agreements of 1958 and which, like them, was accepted by the decision of 1965. Two factors from the provisions in fact show that it may be regarded as having been accepted: the broad terms of Article 1 of the decision whereby Brussels, in addition to Strasbourg and Luxembourg, are to remain the provisional place of work of the Community; and the provisions in Article 12, which state that the decision “shall not affect the provisional places of work of the institutions and departments of the European Communities, as determined by previous decisions of the Governments...”.

As I have already stated (see section 14 supra) those provisions regard the previous agreements as having legal effect. It may thus be inferred that the Parliament is technically justified in permitting its own committees to meet in Brussels. Confirmation of that is to be found in the letter which the President of the Council of Ministers of the Communities sent to the President of the Parliament on 22 October 1977 (Annex 10 to the defence). In it he declared: “It is, furthermore, common knowledge that the parliamentary committees are accustomed to meet in Brussels, with the minimum level of staffing required foiholding their meetings.” The wording is unusually clear. But that is not all: the internal administrative practice of the Parliament confirms that Brussels has always been considered the normal place of work of the committees. That is shown by the fact (referred to at the hearing by counsel for the defendant) that before meeting in Brussels the committees have never submitted to the Bureau a reasoned application as required by Article 10 (3) of the Rules of Procedure when meetings are held away from the seat.

As regards the political groups we may be more categorical. In fact neither the agreements of 1958 nor those of 1965 (and certainly not those of 1981, which are merely confirmatory) deal with them at all. Why? The applicant justifies that silence on the ground of the greater independence which the groups enjoy because of their structure and the role which they perform. I find the explanation persuasive. But it is then necessary to concede that, in such matters at any rate, the Parliament has a certain freedom of decision: a freedom which rests on the institution's power to organize its affairs and on the particular nature of the political groups, whilst it is limited by the agreements in existence concerning the place of work (I have in mind above all the location in Luxembourg of the Parliament's Secretariat and departments, which indirectly affects all the activities of the Parliament) and by the reservation to the Member States of the power to determine the seat. I add — and it is a factor of no small importance — that the practice of holding the meetings of the political groups in Brussels as a rule has never provoked any reaction from the governments.

22. 

Paragraph 3 (c) contains the following decisions: (i) to review “the operation of the Secretariat and technical services... to meet the requirements set out in (a) and (b) above, particularly with a view to avoiding the need for a substantial number of staff of Parliament to travel constantly”; (ii) to make use, with that end in view, “of the latest means of telecommunication both for personal contacts and for document transmission”; (iii) to use “the most advanced techniques... to facilitate cooperation between the institutions, while road, rail and air links between the main centres of activity of the Community must be improved”; (iv) to have the appropriate bodies of the Parliament determine under the guidance of the President and the enlarged Bureau the measures to be taken, to evaluate their costs and before the end of the year to present to the Parliament a report accompanied by appropriate proposals.

It is clear that the second and third subparagraphs present no problems. It cannot be doubted that the Parliament is acting intra vires in reorganizing its own departments by using modern means of telecommunication and any other advanced techniques for facilitating cooperation between the institutions. For the same reasons the fourth subparagraph does not give rise to dispute in so far as it concerns the measures to be adopted in relation to the matters set out in the second and third subparagraphs. However, the same cannot be said of the first subparagraph, which is directly concerned with the Secretariat and the departments of the Parliament.

As I have already indicated, that subparagraph provides that the operation of the Secretariat and technical services must be reviewed in order to avoid the need for staff to travel from one seat to another. It is true that the two succeeding subparagraphs set out a series of measures which should reduce such travel; but it does not appear to me reasonable to interpret the provision in question in the light of the succeeding subparagraphs. In other words, I do not consider that the Parliament decided to overcome the abovementioned difficulties merely by availing itself of more modern telecommunications and more advanced technology. The proposition contained in the first subparagraph is of a general nature (so that in the decision it was given separate treatment); in my opinion, therefore it involves a transfer of staff from Luxembourg to other places, in particular to Brussels.

That gives rise in essence to two problems. The principal question to be considered is whether the contested decision is : (a) merely preparatory, in the sense that it forms part of a procedure to be implemented in a number of stages of which only the last produces legal effects; or (b) independently capable of producing external legal effects and thus subject to judicial review. The second problem, which is logically subordinate to the first, consists in establishing whether, or subject to what limits, the decision, in so far as it concerns the posting of staff, is compatible with the agreements between the governments on this matter.

23. 

With regard to the first point the better view appears to me that the contested resolution constitutes a measure producing legal effects which have repercussions beyond the confines of the institution. In fact by adopting the resolution the Parliament has decided — I repeat “decided”, not merely considered — the transfer of staff from Luxembourg to other places, in particular Brussels. The administration was charged with determining (it must again be emphasized: “determining”, not merely “considering”) the measures necessary for that purpose and their costs, which, in my view, constitutes the implementation of a decision which has already been taken. Nor must we be misled by the fact that the number or the names of the staff to be transferred are not given and that Brussels is not expressly mentioned as the place of destination; these lacunas appear reasonable if it is considered that the Parliament is giving instructions of a general nature and that the administrative authorities are required to take specific measures to implement them. The same applies with regard to the report (mentioned in the fourth subparagraph) that the appropriate bodies must present to the Parliament: that and the subsequent decisions by the institution form part of a specific choice which has already been made definitively in the resolution in question. An enlightening factor in this matter appears in the sixteenth recital in the preamble, in which it is noted that the “Parliament has always asserted its right to meet and work where it chooses”.

24. 

Turning then to the second problem, I will remind the Court that the decision of 8 April 1965, Article 4 of which I have already quoted, was adopted by the governments in exercise of their power to determine the seats of the institutions. Clearly, any measure of the Parliament which involves posting staff from Luxembourg to other places must be appraised in the light of Article 4. However, that provision must not in my view, be interpreted inflexibly, because, amongst other reasons, the Parliament is no longer what it was in 1965 but has been made more representative and has seen its membership doubled.

In order to identify the scope of the provision it is, in any case, indispensable to place it in the context of the agreements, express or tacit, concerning the location of the sessions in Strasbourg and the committees in Brussels. If these agreements are borne in mind everything will appear clear, because it is impossible to see any sense in them and to permit the Parliament — the Parliament of 1982, I stress — to implement them effectively without considering lawful the posting in moderation of staff outside Luxembourg. That was expressly recognized by the President of the Council of Ministers in the letter of 22 September 1977 which I have already quoted. I therefore consider that, on the basis of the agreements between the governments the Parliament is empowered to transfer a certain number of officials to Brussels, provided that they constitute staff indispensable for the provision in situ of the minimum facilities necessary for the work of the committees. I naturally rule out that such transfers may concern whole departments or offices.

In order to develop a criterion which is both practicable and fair it may be of assistance to consider the present distribution of personnel between Luxembourg and Brussels. A detailed picture of that distribution is provided by Annex 2 to the Parliament's reply to the questions submitted to it by the Court (see the document lodged at the Registry on 15 June 1982). In 1982 the position was as follows: out of 2612 established officials 117 serve in Brussels; in Brussels there are also 155 temporary servants assigned to the political groups. I consider that if the transfer to Brussels of staff (established officials or other servants) were to bring about a significant alteration in the present proportions of staff assigned to Luxembourg and Brussels it would constitute a measure of internal organization clearly at variance with the agreements between the governments on the location of the institutions and accordingly vitiated by lack of competence.

In that state of affairs, in order to hold the resolution of the Parliament vitiated by lack of competence we should need to establish: (a) that it affects the distribution of staff amongst the various centres where the institution works; and (b) that the posting of staff from Luxembourg to other places envisaged in the resolution is of such a nature as to conflict with Article 4 of the 1965 decision interpreted in the manner which I have suggested above. I do not consider that a conclusion of that nature is justified. I am led to take this view above all by the nonspecific nature of the resolution with regard to the transfer of staff. The subject-matter — which is delicate and indeed thorny — has clearly led the Parliament to be careful in what it says and leaves unsaid. The most appropriate course is accordingly to propose that the Court should declare the measure lawful in so far as it Ís interpreted and implemented in accordance with Article 4 of the 1965 decision.

25. 

On the basis of all the considerations which I have hitherto set out, I propose that the Court, in deciding the action brought by the Grand Duchy of Luxembourg against the Parliament of the European Communities by an application lodged at the Registry on 7 August 1981, should:

With regard to the costs, it seems equitable to me that the parties should bear their own costs in view of the complexity and novelty of the questions at issue.