United States v. Louisiana (Louisiana Boundary Case)/Opinion of the Court

In United States v. Louisiana, 363 U.S. 1, 80 S.Ct. 961, 4 L.Ed.2d 1025, the Court held that by the Submerged Lands Act of 1953 the United States had quitclaimed to Louisiana the lands underlying the Gulf of Mexico within three geographical miles of the coastline. The United States was declared entitled to the lands further seaward. In the decree, as in the Submerged Lands Act, 'coast line' was defined as 'the line of ordinary low water along that portion of the coast which is in direct contact with the open sea and the line marking the seaward limit of inland waters.' We reserved jurisdiction 'to entertain such further proceedings, enter such orders and issue such writs as may * *  * be deemed necessary or advisable to give proper force and effect to this decree.' Before the Court now are cross-motions by the United States and Louisiana for a supplemental decree designating the boundary of the lands under the Gulf owned by Louisiana. The segments of that boundary line that lie three miles outward from 'that portion of the coast which is in direct contact with the open sea' are for the most part easily determinable. The controversy here is primarily over the location of that part of the coastline that consists of 'the line marking the seaward limit of inland waters.'

More than three years ago, in United States v. California, 381 U.S. 139, 85 S.Ct. 1401, 14 L.Ed.2d 296, we held that Congress had left to the Court the task of defining 'inland waters' and we adopted for purposes of the Submerged Lands Act the definitions contained in the International Convention on the Territorial Sea and the Contiguous Zone, ratified by the United States in 1961. The United States asserts that the same definitions should determine the location of the 'line marking the seaward limit of inland waters' of Louisiana. Louisiana, on the other hand, contends that this line has already been determined pursuant to an 1895 Act of Congress which directed the drawing of 'lines dividing the high seas from rivers, harbors and inland waters,' and has proposed a decree based upon this contention. Alternatively, Louisiana argues that, even assuming the applicability of the definitions contained in the Convention on the Territorial Sea and the Contiguous Zone, the decree proposed by the United States reflects too restrictive a construction of the Convention's provisions in derogation of relevant principles of international law.

The 'Inland Water Line.'

Comprehensive congressional regulation of maritime navigation began with the Act of April 29, 1864, which promulgated rules applicable to all vessels of domestic registry on any waters. These rules were patterned on emerging international standards, and when most other maritime nations subsequently changed their rules, the United States Congress in 1885 enacted conforming 'Revised International Rules and Regulations' to govern American ships 'upon the high seas and in all coast waters of the United States, except such as are otherwise provided for.' The 1864 Act was therefore repealed except as to navigation 'within the harbors, lakes, and inland waters of the United States.' In 1889 the International Maritime Conference drafted new International Rules, which were promptly adopted by Congress. Article 30 of those rules provided that '(n)othing in these rules shall interfere with the operation of a special rule, duly made by local authority, relative to the navigation of any harbor, river, or inland waters.'

The United States already had in the 1864 Act such special inland rules for ships of American registry. In order to clarify the areas and ships to which the International and Inland Rules would respectively apply, Congress in 1895 provided that the rules of the 1864 Act were to govern the navigation of all vessels 'on the harbors, rivers and inland waters of the United States.' The 1895 Act went on to provide:

'The Secretary of the Treasury is hereby authorized,     empowered and directed from time to time to designate and      define by suitable bearings or ranges with light houses,      light vessels, buoys or coast objects, the lines dividing the      high seas from rivers, harbors and inland waters.'

The authority thus vested in the Secretary of the Treasury has since been transferred several times to various federal officials and now resides with the Commandant of the Coast Guard; and from time to time the lines authorized by the 1895 Act have been designated along portions of the United States coast. When the Submerged Lands Act was passed in 1953, such lines had been drawn in the Gulf only along some segments of the Louisiana shore, but in that year the Commandant of the Coast Guard drew new lines applicable to all the waters off the Louisiana coast. In 1954 the Louisiana Legislature declared that it 'accepted and approved' this demarcation, which it now calls the 'Inland Water Line,' as its boundary. Louisiana now argues that this line encloses inland waters and is therefore 'the line marking the seaward limit of inland waters,' and thus its 'coastline' within the meaning of the Submerged Lands Act.

Louisiana argues initially that the 1895 Act is in pari materia with the Submerged Lands Act. Congress, it is said, must have contemplated that a technical term such as 'inland waters' should have the same meaning in different statutes. The phrase appears, however, in quite different contexts in the two pieces of legislation. While the Submerged Lands Act established boundaries between the lands of the States and the Nation, Congress' only concern in the 1895 Act was with the problem of navigation in waters close to this Nation's shores. There is no evidence in the legislative history that it was the purpose of Congress in 1953 to tie the meaning of the phrase 'inland waters' to the 1895 statute. For instance, during the Senate Committee hearings on the Submerged Lands Act, the following exchange took place between Senator Anderson and the Assistant Attorney General of Louisiana:

'Senator ANDERSON. Was there not a so-called Government line     drawn along the coast of Louisiana?

'Mr. MADDEN. Only a partial line, Senator. I remember the old     statute that authorized, I believe it was first assistant      Attorney General of Louisiana: Treasury, to fix a line to      show the demarcation between inland waters and the high seas. I think the Coast Guard has attempted to draw a partial line     over on the east side of Louisiana.

'Senator ANDERSON. We went through all that in the hearing a     couple of years ago, and found that was of no value to us      whatsoever.'

Louisiana's position that the Submerged Lands Act must necessarily be read as referring to the 1895 Act is thus not tenable. After a lengthy review of the legislative history of the Submerged Lands Act in United States v. California, we reached the conclusion that Congress deliberately 'chose to leave the definition of inland waters where it found it-in the Court's hands.' 381 U.S. at 157, 85 S.Ct. at 1412. We adhere to that view, and turn to Louisiana's other arguments in support of the 'Inland Water Line.'

We further decided in United States v. California that the provisions of the Convention on the Territorial Sea and the Contiguous Zone were 'the best and most workable definitions available,' 381 U.S., at 165, 85 S.Ct., at 1415, and we adopted them for purposes of the Submerged Lands Act. Yet Louisiana asserts that the Court is not precluded by the California decision from adopting the 'Inland Water Line' in this case. Essentially the argument is that the Convention was not intended either to be the exclusive determinant of inland or territorial waters or to divest a nation of waters which it had long considered subject to its sole jurisdiction. By the long-standing, continuous, and unopposed exercise of jurisdiction to regulate navigation on waters within the 'Inland Water Line,' the United States is said to have established them as its inland waters under traditional principles of international law. Alternatively, Louisiana suggests that, even assuming the exclusivity of the Convention on the Territorial Sea and the Contiguous Zone, the 'Inland Water Line,' by virtue of this assertion of sovereignty, has created 'historic bays' within the exception of Article 7 of the Convention. We have concluded, however, that nothing in either the enactment of the 1895 Act or its administration indicates that the United States has ever treated that line as a territorial boundary.

Under generally accepted principles of international law, the navigable sea is divided into three zones, distinguished by the nature of the control which the contiguous nation can exercise over them. Nearest to the nation's shores are its inland, or internal waters. These are subject to the complete sovereignty of the nation, as much as if they were a part of its land territory, and the coastal nation has the privilege even to exclude foreign vessels altogether. Beyond the inland waters, and measured from their seaward edge, is a belt known as the marginal, or territorial sea. Within it the coastal nation may exercise extensive control but cannot deny the right of innocent passage to foreign nations. Outside the territorial sea are the high seas, which are international waters not subject to the dominion of any single nation.

Whether particular waters are inland has depended on historical as well as geographical factors. Certain shoreline configurations have been deemed to confine bodies of water, such as bays, which are necessarily inland. But it has also been recognized that other areas of water closely connected to the shore, although they do not meet any precise geographical test may have achieved the status of inland waters by the manner in which they have been treated by the coastal nation. As we said in United States v. California, it is generally agreed that historic title can be claimed only when the 'coastal nation has traditionally asserted and maintained dominion with the acquiescence of foreign nations.' 381 U.S., at 172, 85 S.Ct. at 1419.

While there is not complete accord on the definition of historic inland waters, it is universally agreed that the reasonable regulation of navigation is not alone a sufficient exercise of dominion to constitute a claim to historic inland waters. On the contrary, control of navigation has long been recognized as an incident of the coastal nation's jurisdiction over the territorial sea. Article 17 of the Convention on the Territorial Sea and the Contiguous Zone embodies this principle in its declaration that '(f)oreign ships exercising the right of innocent passage (in the territorial sea) shall comply with the laws and regulations enacted by the coastal State * *  * and, in particular, with such laws and regulations relating to transport and navigation.' Because it is an accepted regulation of the territorial sea itself, enforcement of navigation rules by the coastal nation could not constitute a claim to inland waters from whose seaward border the territorial sea is measured.

But even if a nation could base a claim to historic inland waters on its continuous regulation of navigation, it is clear that no historic title can accrue when the coastal nation disclaims any territorial reach by such an exercise of jurisdiction. For at leas the last 25 years, during which time Congress has twice re-enacted both the International and Inland Rules, the responsible officials have consistently disclaimed any but navigational significance to the 'Inland Water Line.' When the line was for the first time completed off the entire Louisiana shore, the Commandant of the Coast Guard declared:

'The establishment of descriptive lines of demarcation is     solely for purposes connected with navigation and shipping. *     *  * These lines are not for the purpose of defining Federal      or State boundaries, nor do they define or describe Federal      or State jurisdiction over navigable waters.'

As early as 1943 the Coast Guard had differentiated the 'Inland Water Line' from other boundaries with territorial significance. Its manual on Admiralty Law Enforcement, published that year, discussed the principles of international law relating to the definitions and jurisdictional attributes of inland waters, the territorial sea, and the high seas. The manual then contrasted the line drawn under the 1895 Act.

'NAVIGATION RULE: Now let us consider another line of     demarcation. As shown in Chapter V, there are different rules     for navigation on the 'inland waters' and the 'high seas':      the Inland Rules and the International Rules. But here we do not apply the previous definition, but adopt a new one for     convenience. The Secretary of Commerce has fixed a series of     lines along our coast, lines not following the natural      curvature of our shores, and not following any three-mile      natural perimeter, and the Inland Rules apply inside this      line, while the International Rules apply outside the line. *     *  *

'Quite obviously, this artificial line does not truly     separate the high seas from the inland waters of the United      States. It simply marks the area within which the Inland     Rules apply, and outside of which the International Rules      control.'

In United States v. California we held that the United States' disclaimer to the Court of any historic title was decisive in the light of the 'questionable evidence of continuous and exclusive assertions of dominion over the disputed waters.' 381 U.S., at 175, 85 S.Ct., at 1421. In this case, not only are there long-standing, extrajudicial disclaimers of historic title, but also the United States as never treated the 'Inland Water Line' as delimiting an area within which it can exercise jurisdiction over anything but navigation.

There is no indication that in enacting the navigation rules and authorizing the designation of an 'Inland Water Line' Congress believed it was also determining the Nation's territorial boundaries. Indeed, it seems unlikely that Congress, if it had intended that result would have delegated such authority to the Secretary of the Treasury, to be exercised in his discretion 'from time to time' and by reference to navigational aids rather than in accordance with prevailing principles of international law. Consistently with their limited statutory purpose, the lines have always been drawn, and frequently altered, solely with regard to contemporary navigational needs. And in the only instance called to our attention in which the 'Inland Water Line' was mentioned by the United States in its international relations, the State Department in 1929 cautioned that the 'lines do not represent territorial boundaries, but are for navigational purposes.' We must therefore reject Louisiana's contention that the United States has historically treated the 'Inland Water Line' as the territorial boundary of its inland waters.

Finally, Louisiana argues that only adoption of the current 'Inland Water Line' will fulfill the 'requirements of definiteness and stability which should attend any congressional grant of property rights belonging to the United States.' United States v. California, 381 U.S. 139, 167, 85 S.Ct. 1401, 1416, 14 L.Ed.2d 296. Any line drawn by application of the rules of the Convention on the Territorial Sea and the Contiguous Zone would be ambulatory and would vary with the frequent changes in the shoreline. This will lead, it is said, to continuing uncertainty and endless litigation concerning the location of the Louisiana coastline under the Submerged Lands Act, because the shoreline is constantly shifting as the Mississippi River and violent Gulf storms remold the soft, silt-like delta soil. This problem was not encountered on the rockhard, comparatively straight California coast, and Louisiana contends that there is nothing in the Submerged Lands Act which requires that inland waters be given the same definition for every part of the United States coast. Just as the Court was free in United States v. California to adopt the definition which best solved the problems of that case, the argument concludes, we are free in this case to adopt a different definition more suited to the peculiarities of the highly unstable Louisiana shore.

We do not, however, so broadly construe our function under the Submerged Lands Act. Our adoption in United States v. California of the definitions contained in the Convention on the Territorial Sea and the Contiguous Zone was 'for purposes of the Submerged Lands Act,' and not simply for the purpose of delineating the California coastline. Congress left to this Court the task of defining a term used in the Act, not of drawing state boundaries by whatever method might seem appropriate in a particular case. It would be an extraordinary principle of construction that would authorize or permit a court to give the same statute wholly different meanings in different cases, and it would require a stronger showing of congression l intent than has been made in this case to justify the assumption of such unconfined power. Finally, we note that if the inconvenience of an ambulatory coastline proves to be substantial, there is nothing in this decision which would obstruct resolution of the problems through appropriate legislation or agreement between the parties. Such legislation or agreement might, for example, freeze the coastline as of an agreed-upon date.

Even if we were free to adopt varying definitions of inland waters for different portions of the United States coast, we are not convinced that the policy in favor of a certain and stable coastline, strong as it is, would necessarily outweigh countervailing policy considerations under the Submerged Lands Act. We recognized in California the desirability of 'a single coastline for both the administration of the Submerged Lands Act and the conduct of our future international relations.' 381 U.S., at 165, 85 S.Ct., at 1415. The adoption of the 'Inland Water Line' for Louisiana would be completely at odds with this desideratum. Moreover, adoption of a new definition of inland waters in this case would create uncertainty and encourage controversy over the coastlines of other States, unsure as to which, if either, of the two definitions would be applied to them. This uncertainty might be compounded by the absence of any 'Inland Water Line' around much of the United States. And we cannot assume that, in enacting the Submerged Lands Act, Congress envisioned that the ownership of potentially vast resources might thereafter be determined 'from time to time' by the Coast Guard, acting solely in the interest of navigational convenience.

For these reasons, we conclude that that part of Louisiana's coastline which, under the Submerged Lands Act, consists of 'the line marking the seaward limit of inland waters,' is to be drawn in accordance with the definitions of the Convention on the Territorial Sea and the Contiguous Zone.

APPLICATION OF THE CONVENTION ON THE TERRITORIAL SEA AND THE CONTIGUOUS ZONE.

Many issues divide the parties concerning the application of the provisions of the Convention on the Territorial Sea and the Contiguous Zone to the Louisiana coast. Some of these issues, which involve simply interpretation of the Convention, we have been able to decide on the basis of the materials now before us. Others, however, are primarily factual questions involving the construction and application of the Convention's provisions with respect to particularized geographical configurations. Several of these factual disputes cannot be properly resolved without evidentiary hearings, and as to others we think it would be wise at all events in this technical and unfamiliar area to have the benefit, preliminarily, of the judgment of a detached referee. Accordingly, we have decided to refer to a Special Master the task of resolving in the first instance several of the particularized disputes over the precise boundary between the submerged Gulf lands belonging to the United States and those belonging to Louisiana.

1. Dredged channels. A recurring question in the application of the Convention to the Louisiana coast is whether dredged channels in the Gulf leading to inland harbors comprise inland waters. In support of its contention that dredged channels, as such, are inland waters, Louisiana relies principally on Article 8 of the Convention:

'For the purpose of delimiting the territorial sea, the     outermost permanent harbour works which form an integral part      of the harbour system shall be regarded as forming part of      the coast.'

Incontestably, Louisiana argues, the channels 'form an integral part of the harbour system'; that they are 'harbour works' as well should also be obvious in light of the enormous cost and effort which the United States has expended in dredging and maintaining them.

The United States argues more convincingly, however, that Article 8 applies only to raised structures. The discussions of the Article by the 1958 Geneva Conference and the International Law Commission reveal that the term 'harbour works' connoted 'structures' and 'installations' which were 'part of the land' and which in some sense enclosed and sheltered the waters within. It is not enough that the dredged channels may be an 'integral part of the harbour system'; even raised structures which fit that description, such as lighthouses, are not considered 'harbour works' unless they are 'connected with the coast.' Thus, Article 8 provides that 'harbour works * *  * shall be regarded as forming part of the coast' (emphasis supplied), a description which hardly fits underwater channels. As part of the 'coast,' the breadth of the territorial sea is measured from the harbor works' low-water lines, attributes not possessed by dredged channels. We must therefore conclude that Article 8 does not establish dredged channels as inland waters.

Louisiana also contends that the legislative history of the Submerged Lands Act reveals a clear congressional purpose to include such channels as inland waters. Early versions of the bill contained a definition of the term 'inland waters' for the purposes of the Act, and that definition included 'channels.' The definition was later deleted, but Louisiana contends that the sole purpose of the deletion was to avoid a construction of the definition which would exclude other areas from inland waters. In United States v. California, 381 U.S. 139, 150-160, 85 S.Ct. 1401, 1407-1413, 14 L.Ed.2d 296, we reviewed at length the pertinent legislative history and concluded that the only sure inference which could be drawn from the deletion of the definition was that Congress thought the highly technical question should be left to the courts. We remain of that view. Moreover, it is far from clear that the word 'channels' in the deleted definition encompassed dredged channels in the open sea. From the context in which the word appears, it is far more likely that the definition referred only to bodies of water bordered by land.

2. The territorial sea of low-tide elevations. Article 11 of the Convention on the Territorial Sea and the Contiguous Zone deals with the subject of low-tide elevations:

'1. A low-tide elevation is a naturally-formed area of land     which is surrounded by and above water at low-tide but      submerged at high tide. Where a low-tide elevation is     situated wholly or partly at a distance not exceeding the      breadth of the territorial sea from the mainland or an      island, the low-water line on that elevation may be used as      the baseline for measuring the breadth of the territorial      sea.

'2. Where a low-tide elevation is wholly situated at a     distance exceeding the breadth of the territorial sea from      the mainland or an island, it has no territorial sea of its      own.'

The question presented by the application of this provision to the Louisiana coast is whether the territorial sea-or, for purposes of this case, the three-mile grant to Louisiana under the Submerged Lands Act-is to be measured from low-tide elevations which lie within three miles of the baseline across the mouth of a bay but more than three miles from any point on the mainland or an island.

The United States argues that the phrase 'at a distance not exceeding the breadth of the territorial sea from the mainland' does not refer to the territorial sea as a situs. Rather it uses the width of the territorial sea only as a measurement of distance a circumlocution made necessary by the failure of the 1958 Geneva Conference to agree upon a uniform width. And that distance-three miles in this case-is to be measured from the 'mainland,' a term which does not comprise baselines across bodies of water but is limited to the low-water mark on dry land. Louisiana, on the other hand, interprets the Article as covering all low-tide ele ations situated anywhere within the territorial sea. And the drawing of baselines across the mouths of bays is an integral step in the determination of the area of the territorial sea. Moreover, Louisiana argues, the term 'mainland' does include inland waters. The theory of the Convention, it is argued, reflects a longstanding principle of international law-that bays and other inland waters are practically assimilated to the dry land and treated for all legal purposes as if they were a part of it.

The parties agree that Article 11 on its face is not wholly dispositive of the issue, and that the language does not preclude either construction. Each party, therefore, relies on the origins of the Article and the statements of its drafters. When the provision was first proposed to the International Law Commission in 1952, it read as follows:

'Elevations of the sea bed situated within the territorial     sea, though only above water at low tide, are taken into consideration for the determination of the      base line of the territorial sea.' (Emphasis supplied.)

After several amendments to the rapporteur's draft, the Commission in 1954 adopted a version with substantially the same meaning:

'Drying rocks and shoals which are wholly or partly within     the territorial sea may be taken as points of departure for      delimiting the territorial sea.' (Emphasis supplied).

As the discussion made clear, both drafts of the Article covered all low-tide elevations within the territorial sea, however measured. Moreover, the provision was thought to embody long-standing principles of international law.

The draft encountered a serious objection, however, which led to its further amendment by the International Law Commission. If every low-tide elevation 'within the territorial sea' was to have a territorial sea of its own, then

'a country like Holland might extend its territorial sea very     considerably by advancing from one shoal to another, claiming      that a shoal situated within the territorial sea of another      shoal had itself a territorial sea.'

To avoid this undue extension of the territorial sea, the final draft of the Commission was revised to read as follows:

'Drying rock and drying shoals which are wholly or partly     within the territorial sea, as measured from the mainland or      an island, may be taken as points of departure for measuring      the extension of the territorial sea.' (Emphasis supplied.)

It is clear that under the International Law Commission version of Article 11, the 'territorial sea, as measured from the mainland' included those portions which extended from baselines enclosing bays. The sole purpose of the amendment to the initial proposals was to indicate that 'drying rocks and drying shoals could only be used once as points of departure for extending the territorial sea and that the process could not be repeated by leap frogging, as it were, from one rock or shoal to another.'

The United States contends that by changing the language of the International Law Commission draft to its present form in the Convention, the Geneva Conference intended also to change its meaning. Precisely the opposite conclusion, however, flows from an inspection of the history of the Convention. The amendment was advanced by the United States; yet its explanation for the proposal contained not the slightest indication that any change in the basic meaning of the Article was intended. Surely there would have been some discussion of the reference to the territorial sea as a measure of distance rather than as a situs had it been the purpose of the United States or the Conference to alter so significantly the meaning of prior drafts and the existing international consensus. Instead, the expert to the Secretariat of the Conference explained 'that all the proposals on article 11 corresponded entirely to the intentions of the International Law Commission.' We therefore conclude that low-tide elevations situated in the territorial sea as measured from bay-closing lines are part of the coastline from which the three-mile grant of the Submerged Lands Act extends.

3. The semicircle test. Article 7(2) defines a bay as follows:

'For the purposes of these articles, a bay is a well-marked     indentation whose penetration is in such proportion to the      width of its mouth as to contain landlocked waters and      constitute more than a mere curvature of the coast. An     indentation shall not, however, be regarded as a bay unless      its area is as large as, or larger than, that of the      semi-circle whose diameter is a line drawn across the mouth      of that indentation.'

(a) In several areas along the Louisiana coast the parties raise the problem of whether and to what extent indentations within or tributary to another indentation can be included in the area of the latter for purposes of the semicircle test. Louisiana argues that a closing line should be drawn across what it calls 'Outer Vermilion Bay' from Tigre Point to Shell Keys. That body of water does not meet the semicircle test unless the area of Vermilion Bay, joined to 'Outer Vermilion Bay' only by a channel between the mainland and Marsh Island, is included. Similarly Louisiana contends that 'Ascension Bay,' whose headlands are said to be the jetties at Belle Pass on the west and Southwest Pass on the east, is a bay under Article 7(2). Again, however, its area will satisfy the semi-circle test only if deemed to include the waters of the Barataria Bay-Caminada Bay complex, which are separated from the outer indentation by a string of islands.

Louisiana argues that the area of tributary bays or other indentations must be included within that of the primary indentation. Article 7(3) provides that '(f)or the purpose of measurement, the area of an indentation is that lying between the low-water mark around the shore of the indentation and a line joining the low-water marks of its natural entrance points.' (Emphasis supplied.) The italicized phrase, it is said, constitutes a direction to follow the low-water line wherever it goes, including into other indentations, in drawing the perimeter of the primary bay. The general rule is well recognized, Louisiana argues, by the United States Department of State among others, that the area of bays within bays is included in calculating the semicircle test.

The United States does not reject the notion that some indentations which would qualify independently as bays may nonetheless be considered as part of larger indentations for purposes of the semicircle test; but it denies the existence of any rule that all tributary waters are so includible. Article 7(2), it emphasizes, refers to 'that indent tion.' The inner bays can be included, therefore, only if they can reasonably be considered part of the single, outer indentation. And that cannot be said of inland waters which, like Vermilion Bay and Barataria Bay-Caminada Bay, are wholly separated from the outer body of water and linked only by narrow passages or channels.

For purposes of this lawsuit, we find it unnecessary to provide a complete answer to the questions posed by the parties. 'Outer Vermilion Bay,' if it is to qualify under the semicircle test, must include the waters of Vermilion Bay. Yet Vermilion Bay is itself a part of the much larger indentation which includes West and East Cote Blanche Bays and Atchafalaya Bay, and which opens to the sea between Marsh Island and Point au Fer. Recognition of the unitary nature of this larger indentation follows from Louisiana's insistence that the low-water mark must be followed around the entire indentation. If, as Louisiana posits, the western headland of the indentation is at Tigre Point, then a closing line across its mouth to Point au Fer far exceeds the 24-mile limit imposed by Article 7(4). It follows that 'Outer Vermilion Bay' is neither itself a bay nor part of a larger bay under the Convention on the Territorial Sea and the Contiguous Zone.

We have concluded, on the other hand, that the area of 'Ascension Bay' does include the Barataria Bay-Caminada Bay complex and therefore meets the semicircle test. Those inner bays are separated from the larger 'Ascension Bay' only by the string of islands across their entrances. If those islands are ignored, the entrance to Barataria and Caminada Bays is sufficiently wide that those bays and 'Ascension Bay' can reasonably be deemed a single large indentation even under the United States' approach. Article 7(3) provides that for the purposes of calculating the semicircle test, '(i)slands within an indentation shall be included as if they were part of the water areas of the indentation.'

The clear purpose of the Convention is not to permit islands to defeat the semicircle test by consuming areas of the indentation. We think it consistent with that purpose that islands should not be permitted to defeat the semicircle test by sealing off one part of the indentation from the rest. Treating the string of islands 'as if they were part of the water areas' of the single large indent tion within which they lie, 'Ascension Bay' does meet the semicircle test.

(b) Another issue involving the semicircle test arises in East Bay in the Mississippi River Delta. Since East Bay does not meet the semicircle test on a closing line between its seawardmost headlands-the tip of the jetty at Southwest Pass and the southern end of South Pass-it does not qualify as a bay under Article 7 of the Convention on the Territorial Sea and the Contiguous Zone. There is a line which can be drawn within East Bay, however, so as to satisfy the semicircle test. Louisiana argues that, just as under Article 7(5) a 24-mile line can be drawn within a bay whose mouth is more than 24 miles wide, so also can a line which satisfies the semicircle test be drawn within a bay whose mouth is too wide to meet that test.

The analogy is unsound. A bay whose mouth is wider than 24 miles is nevertheless a bay. But an indentation that does not meet the semicircle test is not a bay but open sea. If an indentation which satisfies the semicircle test is a true bay, therefore, it cannot be on the theory that the closing line carves out a portion of a larger bay. The enclosed indentation must by its own features qualify as a bay.

The United States argues that the area within East Bay enclosed by Louisiana's proposed line does not constitute a bay because there is no 'well-marked indentation' with identifiable headlands which encloses 'landlocked' waters. Indeed, it is said, there is not the slightest curvature of the coast at either asserted entrance point. We do not now decide whether the designated portion of East Bay meets these criteria, but hold only that they must be met. We cannot accept Louisiana's argument that an indentation which satisfies the semicircle test ipso facto qualifies as a bay under the Convention. Such a construction would fly in the face of Article 7(2), which plainly treats the semicircle test as a minimum requirement. And we have found nothing in the history of the Convention which would support so awkward a construction.

4. Islands at the mouth of a bay. Article 7(3) of the Convention on the Territorial Sea and the Contiguous Zone provides:

'For the purpose of measurement, the area of an indentation     is that lying between the low-water mark around the shore of the indentation and a line joining      the low-water marks of its natural entrance points. Where,     because of the presence of islands, an indentation has more      than one mouth, the semi-circle shall be drawn on a line as      long as the sum total of the lengths of the lines across the      different mouths. Islands within an indentation shall be     included as if they were part of the water areas of the      indentation.'

While the only stated relevance of such islands is to the semicircle test, it is clear that the lines across the various mouths are to be the baselines for all purposes. The application of this provision to the string of islands across the openings to the Lake Pelto-Terrebonne Bay-Timbalier Bay complex has raised the following questions: (a) between what points on the islands are the closing lines to be drawn, and (b) should the lines be drawn landward of a direct line between the entrance points on the mainland?

(a) It is Louisiana's primary contention that when islands appear in the mouth of a bay, the lines closing the bay and separating inland from territorial waters should be drawn between the mainland headlands and the seawardmost points on the islands. This position, however, is refuted by the language of Article 7(3), which provides for the drawing of baselines 'across the different mouths' (emphasis supplied), not across the most seaward tips of the islands. There is no suggestion in the Convention that a mouth caused by islands is to be located in a manner any different from a mouth between points on the mainland that is, by 'a line joining the low-water marks of (the bay's) natural entrance points.' The 'natural entrance points' may, and in some instances in the Lake Pelto-Terrebonne Bay-Timbalier Bay complex do, coincide with the outermost edges of the islands. But there is no automatic correlation, and the headlands must be selected according to the same principles that govern the location of entrance points on the mainland.

(b) Louisiana argues in the alternative that even if the closing lines should not necessarily connect the most seaward points on the islands, in no event should they be drawn landward of a direct line between the entrance points on the mainland. The purpose of Article 7(3) is expressed in the following passage from the Commentary of the International Law Commission:

'Here, the Commission's intention was to indicate that the     presence of islands at the mouth of an indentation tends to      link it more closely to the mainland, and this consideration      may justify some alteration in the ratio between the width      and the penetration of the indentation.'

It is evident, Louisiana argues, that Article 7(3) was designed to enlarge rather than contract the area of inland waters; and that this policy would not be served by permitting islands intersected by a direct closing line between the mainland headlands to pull that line inward, particularly when the indentation would qualify as a bay even in the absence of the islands. Rather, the line should be selected which will enclose the maximum area of island waters.

Louisiana's argument is undermined, however, by the natural effect of islands at the mouth of an indentation described in the International Law Commission Commentary. Just as the 'presence of islands at the mouth of an indentation tends to link it more closely to the mainland,' so also do the islands tend to separate the waters within from those without the entrances to the bay. Even waters which would be considered within the bay and therefore 'landlocked' in the absence of the islands are physically excluded from the indentation if they lie seaward of the mouths between the islands. It would be anomalous indeed to say that waters are part of a bay even though they lie outside its natural entrance points. No doubt there could be islands which would not, whether because of their size, shape, or relationship to the mainland, be said to create more than one mouth to the bay. But where, as in the Lake Pelto-Terrebonne Bay-Timbalier Bay complex, a string of islands covers a large percentage of the distance between the mainland entrance points, the openings between the islands are distinct mouths outside of which the waters cannot sensibly be called 'inland.'

Louisiana purports to find support for its position in the provision of Article 7(3) that '(i)lands within an indentation shall be included as if they were part of the water areas of the indentation.' This provision would preclude drawing lines to an island wholly within the indentation, Louisiana argues, and it should therefore also preclude drawing closing lines to any part of an island landward of a straight line between the mainland headlands. We cannot, however, accept this construction of the Convention. An island which is intersected by a direct mainland-to-mainland closing line is not 'within (the) indentation.' Nor can an island which forms the mouth of an indentation be 'within' it.

Article 7(3) clearly distinguishes between islands which, by creating multiple mouths, form a part of the perimeter of the bay, and those which, by their presence wholly 'within' the bay, are treated as part of its water areas.

In sum, we hold that where islands intersected by a direct closing line between the mainland headlands create multiple mouths to a bay, the bay should be closed by lines between the natural entrance points on the islands, even if those points are landward of the direct line between the mainland entrance points.

5. Islands as headlands of bays. With respect to many of the bays on the Louisiana coast the question is presented whether a headland of an indentation can be located on an island. The United States argues that the Convention on the Territorial Sea and the Contiguous Zone flatly prohibits the drawing of bay-closing lines to islands. A true bay, it is said, is an 'indentation' within the mainland, and it cannot be created by the 'projection' of an island or islands from the coast. Moreover, the rule of Article 7(3) that the area of an indentation lies between the closing line and 'the low-water mark around the shore of the indentation' contemplates a perimeter of dry land unbroken by any opening other than the bay's entrance. Finally, the United States argues, such an opening between the island and the mainland would deprive the enclosed waters of the 'landlocked' quality required in a true bay.

We do not agree that the face of the Convention clearly concludes the question. No language in Article 7 or elsewhere positively excludes all islands from the meaning of the 'natural entrance points' to a bay. Waters within an indentation which are 'landlocked' despite the bay's wide entrance surely would not lose that characteristic on account of an additional narrow opening to the sea. That the area of a bay is delimited by the 'low-water mark around the shore' does not necessarily mean that the low-water mark must be continuous.

Moreover, there is nothing in the history of the Convention or of the international law of bays which establishes that a piece of land which is technically an island can never be the headland of a bay. Of course, the general understanding has been-and under the Convention certainly remains that bays are indentations in the mainland, and that islands off the shore are not headlands but at the most create multiple mouths to the bay. In most instances and on most coasts it is no doubt true that islands would play only that restricted role in the delimitation of bays. But much of the Louisiana coast does not fit the usual mold. It is marshy, insubstantial, riddled with canals and other waterways, and in places consists of numerous small clumps of land which are entirely surrounded by water and therefore technically islands. With respect to some spots along the Louisiana coast even the United States has receded from its rigid position and recognized that these insular configurations are really 'part of the mainland.' The western shore of the Lake Pelto-Terrebonne Bay-Timbalier Bay indentation is such a formation, and is treated by the United States as part of the coast.

This Court too has in the past adopted this realistic approach to similar land formations. In Louisiana v. Mississippi, 202 U.S. 1, 45-46, 26 S.Ct. 408, 420, 50 L.Ed. 913, we wrote:

'Mississippi denies that the peninsula of St. Bernard and the     Louisiana marshes constitute a peninsula in the true sense of      the word, but insists that they constitute an archipelago of      islands. Certainly there are in the body of the Louisiana     marshes or St. Bernard peninsula portions of sea marsh which      might technically be called islands, because they are land      entirely surrounded by water, but they are not true islands. They are rather, as the Commissioner of the General Land     Office wrote the Mississippi land commissioner in 1904, 'in      fact, hummocks of land surrounded by the marsh and swamp in      said townships. * *  * '

'And when the Louisiana act used the words: 'thence bounded     by the said gulf to the place of beginning including all      islands within three leagues of the coast,' the coast      referred to is the whole coast of the state, and the      peninsula of St. Bernard formed an integral part of it.' Naturally this common-sense approach extends to coastal formations where there are only a few islands, or even a single island, as well as to those where there are many. Such has been the view of other court and of textwriters.

Much of the Louisiana coast on or near the Mississippi River Delta is of the same general consistency as the western shore of the Lake PeltoTerrebonne Bay-Timbalier Bay complex, and some of the islands may be so closely linked to the mainland as realistically to be assimilated to it. While there is little objective guidance on this question to be found in international law, the question whether a particular island is to be treated as part of the mainland would depend on such factors as its size, its distance from the mainland, the depth and utility of the intervening waters, the shape of the island, and its relationship to the configuration or curvature of the coast. We leave to the Special Master the task of determining in the first instance-in the light of these and any other relev nt criteria and any evidence he finds it helpful to consider-whether the islands which Louisiana has designated as headlands of bays are so integrally related to the mainland that they are realistically parts of the 'coast' within the meaning of the Convention on the Territorial Sea and the Contiguous Zone.

6. Fringes of islands. At several places the question is raised whether areas between the mainland and fringes or chains of islands along the coast are inland waters. The parties agree that no article of the Convention specifically provides that such areas are inland waters. Louisiana argues that they are inland waters, under any one of several theories: that such island fringes form the perimeter of bays under Article 7, that straight baselines must be drawn along the islands under Article 4, or that the waters should be deemed 'inland' under general principles of international law which antedate and supplement the Convention on the Territorial Sea and the Contiguous Zone. The position of the United States is that such island chains can be taken into account as enclosing inland waters only by drawing straight baselines; yet the decision whether the draw such baselines is within the sole discretion of the Federal Government, and the United States has not chosen to do so.

We have concluded that Article 7 does not encompass bays formed in part by islands which cannot realistically be considered part of the mainland. Article 7 defines bays as indentations in the 'coast,' a term which is used in contrast with 'islands' throughout the Convention. Moreover, it is apparent from the face and the history of the Convention that such insular formations were intended to be governed solely by the provision in Article 4 for straight baselines. The language of Article 4 itself is the clearest indication of that intent:

'1. In localities where the coast line is deeply indented and     cut into, or if there is a fringe of islands along the coast      in its immediate vicinity, the method of straight baselines      joining appropriate points may be employed in drawing the      baseline from which the breadth of the territorial sea is      measured.' (Emphasis supplied.)

The drafters of the Convention and their predecessors were aware that international law permitted such island fringes in some circumstances to enclose inland waters. The principle was recognized and applied by the International Court of Justice in the Fisheries Case (United Kingdom v. Norway), (1951) I.C.J. 116, in which Norway was held legitimately to have drawn straight baselines along the 'skjaergaard,' literally a 'rock rampart' composed of hundreds of thousands of insular formations which ringed the mainland. Thereafter, with the Fisheries Case as the model, attempts were made to draft concrete rules for the uniform treatment of such island fringes, and both the International Law Commission and the 1958 Geneva Conference discussed the problem at length. There was, however, too little technical information or consensus among nations on that and related subjects to allow the formulation of uniform rules. It was agreed, therefore, that the problem should be handled as it had been by the International Court of Justice in the Fisheries Case: each nation was left free to draw straight baselines along suitable insular configurations if it so desired. In the light of this resolution of the problem, it is clear that the drafters did not intend to leave island fringes beyond the scope of the Convention altogether. The deliberate decision was that such island formations are not to be treated differently from any other islands unless the coastal nation decides to draw straight baselines.

In United States v. California, 381 U.S. 139, 168, 85 S.Ct. 1401, 1417, 14 L.Ed.2d 296, we held that 'the choice under the Convention to use the straight-base-line method for determining inland waters claimed against other nations is one that rests with the Federal Government, and not with the individual States.' Since the United States asserts that it has not drawn and does not want to draw straight baselines along the Louisiana coast, that disclaimer could, under the California decision, be conclusive of the matter. Louisiana argues, however, that because the Louisiana coast is so perfectly suited to the straight baseline method, and because it is clear that the United States would employ it in the conduct of its international affairs were it not for this lawsuit, the Court should reconsider its holding in California and itself draw appropriate baselines. While we agree that the straight baseline method was designed for precisely such coasts as the Mississippi River Delta area, we adhere to the position that the selection of this optional method of establishing boundaries should be left to the branches of Government responsible for the formulation and implementation of foreign policy. It would be inappropriate for this Court to review or overturn the considered decision of the United States, albeit partially motivated by a domestic concern, not to extend its borders to the furthest extent consonant with international law.

7. Historic inland waters. Louisiana argues that all the waters of the Mississippi River Delta, and East Bay in particular, are 'so-called 'historic' bays' within the meaning of Article 7(6), and that they are therefore inland waters notwithstanding their failure to meet the geographical requirements of Article 7 and the United States' refusal to draw straight baselines. Historic bays are not defined in the Convention, and the term therefore derives its content from general principles of international law. As the absence of a definition indicates, there is no universal accord on the exact meaning of historic waters. There is substantial agreement, however, on the outlines of the doctrine and on the type of showing which a coastal nation must make in order to establish a claim to historic inland waters. But because the concept of historic waters is still relatively impercise and its application to particular areas raises primarily factual questions, we leave to the Special Master-as we did in United States v. California-the task of determining in the first instance whether any of the waters off the Louisiana coast are historic bays. We do not think the ultimate resolution of this litigation would be hastened by any further discussion of the subject at this time, beyond the remarks below.

The Article obviously does not encompass indentations between arms of land formed by the river but not containing it.

In is effort to establish that the waters of the Delta have been subjected to the continuous authority of the coastal nation, Louisiana has relied heavily on is own activities as well as on those of the Federal Government. The United States contends that those state activities cannot in this lawsuit support the position that the Delta waters are historic bays. The argument is not that such exercises of authority by Louisiana would not be relevant to a claim of historic title vis-a -vis another nation. On the contrary, the United States has '(n)o doubt (that) the national government may, if it chooses rely on State action to support its own historic claim as against other nations.' But, the United States asserts, 'a State cannot oblige it to do so or to accept State action as binding in a domestic case such as the present one.' In brief, then, the United States' position is that it can prevent judicial recognition of a ripened claim to historic title merely by lodging a disclaimer with the court.

In United States v. California we noted, but found it unnecessary to pass on, the United States' contention that historic title cannot be founded upon exercises of state authority because a claim to historic inland waters can be maintained only if endorsed by the United States. We there sustained the Master's determination that, even assuming the relevance of California's assertions of sovereignty over the coastal waters, they did not establish historic title. The United States' disclaimer was credited only because the case presented such 'questionable evidence of continuous and exclusive assertions of dominion.' 381 U.S., at 175, 85 S.Ct., at 1421. And we noted that we were 'reluctant to hold that such a disclaimer would be decisive in all circumstances, for a case might arise in which the historic evidence was clear beyond doubt.' Ibid. Thus, the Court indicated its unwillingness to give the United States the same complete discretion to block a claim of historic inland waters as it possesses to decline to draw straight baselines.

While we do not now decide that Louisiana's evidence of historic waters is 'clear beyond doubt,' neither are we in a position to say that it is so 'questionable' that the United States' disclaimer is conclusive. We do decide, however, that the Special Master should consider state exercises of dominion as relevant to the existence of histor c title. The Convention was, of course, designed with an eye to affairs between nations rather than domestic disputes. But, as we suggested in United States v. California, it would be inequitable in adapting the principles of international law to the resolution of a domestic controversy, to permit the National Government to distort those principles, in the name of its power over foreign relations and external affairs, by denying any effect to past events. The only fair way to apply the Convention's recognition of historic bays to this case, then, is to treat the claim of historic waters as if it were being made by the national sovereign and opposed by another nation. To the extent the United States could rely on state activities in advancing such a claim, they are relevant to the determination of the issue in this case.

In due course a Special Master will be appointed by the Court to make a preliminary determination, consistent with this opinion, of the precise boundaries of the submerged lands owned by Louisiana in the Gulf of Mexico. It is so ordered.

Special Master appointed to make preliminary determination.

The CHIEF JUSTICE and Mr. Justice MARSHALL took no part in the consideration or decision of this case. TAI TABLE TAI TABLE

Mr. Justice BLACK, with whom Mr. Justice DOUGLAS joins, dissenting.