FIG. 5. The American proposal regarding bays and estuaries: (a) Since the area between the envelopes of the arcs of circles and the straight headland-to-headland line (shaded in the diagram) exceeds the area of the semicircle (marked "m"), the waters of the bay are interior or national waters, and the straight line becomes the landward boundary of the territorial sea; (b) since the area between the envelopes of the arcs of circles and the straight headland-to-headland line (shaded) is less than the area of the semicircle (marked "m"), the waters of the bay are not interior waters, and the territorial sea is delimited by means of the envelope of the arcs of circles of three-mile radius drawn from all points on the coast.

FIG. 6. The American proposal for the elimination of objectionable pockets of the high sea: The envelope of the arcs of circles of three-mile radius is first drawn from all coasts, both mainland and islands. Where there is a pronounced pocket of high sea which may be wholly enclosed by drawing a single straight line not more than four miles long, such a line is drawn where the entrance first narrows to four miles. If the area between the straight line and the "envelope" (the three shaded areas in the diagram) exceeds the area of a semicircle drawn on the four-mile line (as in the upper and lower shaded areas), the pocket of high sea may be assimilated to the territorial sea. If the area is less than that of the semicircle (as in the middle shaded area) the pocket remains a portion of the high sea.

incide, on any two coasts, in their relative distance and direction from one another, the "envelopes" would be identical, and one could be superposed upon the other. An appreciation of this fact is essential to a thorough understanding of the problems of delimiting the territorial sea.

Since coast lines are of infinite variety, any single rule, no matter how simple, must operate to produce some stretches of territorial sea boundary which will be found unsatisfactory to navigators, to fishermen, or to both. In attempting to avoid or eliminate such undesirable features there are but two alternatives:

(1) A series of rules or principles, each of which is adapted to a particular type of coast;

(2) A single rule, such as the envelope of the arcs of circles, with provision for the elimination of occasional impracticable results.

The first method appears to have been in the back of the minds of the majority of those who have hitherto made studies regarding the delimitation of the territorial sea. This accounts for the proposal to draw straight lines in front of concave coasts, while drawing arcs of circles in front of convex coasts. It accounts also for the efforts which were made by the preparatory committee for the Hague Conference, and at the Conference, to provide still different methods for the delimitation of the territorial sea around the islands of an archipelago. It involves, however, the definition of such terms as "bay" and

"group of islands," and the classification of coast lines into discrete types. This is exceedingly difficult, if not impossible, and almost futile when it is remembered that coast lines of very dissimilar types may have territorial sea boundaries which are strikingly similar if not identical.

The alternative method is the result of studying all aspects of the problem with relation to each other, keeping in mind particularly the point of view of the man on the sea, especially the navigator. This method is to derive the limit of territorial waters from all points of the coast line by a single, simple geometrical procedure, and then to eliminate any impracticable portions of the line by an equally simple and impartial geometrical principle.

It will be found that nearly all of these impracticable results really belong to a single type, namely, small pockets of the high sea which are almost wholly surrounded by territorial waters. The occurrence of such pockets of high sea in bays has been the occasion of much difficulty and serious study. The similarity of the problems created by the occurrence of such pockets of high sea in the vicinity of islands appears not to have been noted.

It has been remarked that there is quite general agreement on the principle of the "envelopes of the arcs of circles" in delimiting territorial waters in front of convex coasts, and that difficulties arise only with respect to concave and highly complicated coast lines. These difficulties relate chiefly to the interior or landward base lines from which the envelopes of the arcs of circles should be drawn.

Assuming the "envelope" method as the general rule of delimiting the territorial sea, two questions arise, both of which relate to the base line from which the three-mile belt is to be measured on certain types of coasts:

(1) What is to be regarded as "land," having a coast line from which to measure at low-water mark, in the case of rocks and shoals awash only at low tide and therefore not constituting bona fide islands?

(2) What are the boundaries between national or inland waters and the territorial sea, which are to be followed in lieu of the coast line in front of pronounced indentations of the coast, such as bays and estuaries?

The first question is one of definition only. The answer does not affect the method of delimiting territorial waters. It matters not whether an "island" is defined as being capable or incapable of use, or as exposed at all stages of the tide or only at low tide. Any such definition simply determines what "land" is to be disregarded altogether in the delimitation of territorial waters. Discussion of this question is omitted from the present study.

The second question involves, in a sense, both definition and delimitation. It is discussed below under the heading "Bays and Estuaries."

BAYS AND ESTUARIES

There is no other aspect of the problem of delimiting territorial waters which has occasioned as much difficulty as the determination of the particular indentations of the coast-whether called bays, gulfs, estuaries, or anything else-whose waters constitute national or interior waters rather than territorial waters. The North Atlantic Fisheries Arbitration Tribunal, for example, decided that

"In case of bays, the three marine miles are to be measured from a straight line drawn across the body of water at the place where it ceases to have the configuration and characteristics of a bay."

7

There is as yet, however, no established rule by which to determine what bodies of water "have the configuration and characteristics of a bay." It is admitted that when an indentation of the coast is regarded as a bona fide bay, it ceases to have the configuration of a bay at its outer headlands.

Assuming that an indentation of the coast constitutes a true bay, there is a question as to its size. In the case of a large gulf the boundary of territorial waters is measured from the sinuosities of the coast. The waters of a small bay, which does not exceed six nautical miles in width, are automatically enclosed by the three-mile rule. A ten-mile width has been frequently accepted, however. The reasons for the ten-mile rule have been well stated by Judge John Bassett Moore:

"Since you observe [he wrote to Mr. Barclay] that there does not appear to be any convincing reason to prefer the ten-mile line in such a case to that of double three miles, I may say that there have been supposed to exist

I Proceedings, p. 97.

8 In a letter quoted in 13 (1894-95) Annuaire de l'Institut de Droit Int., p. 146; quoted in Jessup, The Law of Territorial Waters and Maritime Jurisdiction, p. 356.

The ten-mile line has been reasons both of convenience and of safety. adopted in the cases referred to, as I understand them, as a practical rule. The transgression of an encroachment upon territorial waters by fishing vessels is generally a grave offense, involving in many instances the forfeiture of the offending vessel and it is obvious that the narrower the space in which it is permissible to fish the more likely the offence is to be committed. In order therefore that fishing may be both practicable and safe and not constantly attended with the risk of violating territorial waters, it has been thought to be expedient not to allow it where the extent of free waters, beween the three-mile line drawn on each side of the bay, is less than four miles. This is the reason of the ten-mile line. Its intention is not to hamper or restrict the right to fish, but to render its exercise practicable and safe. When fishermen fall in with a shoal of fish, the impulse to follow it is so strong as to make the possibility of transgression very serious within narrow limits of free water. Hence it has been deemed wiser to exclude them from spaces less than four miles each way from the forbidden lines. In spaces less than this, operations are not only hazardous, but so circumscribed as to render them of little practical value."

The report of the Commission on Territorial Waters, at the recent Hague Conference, contains the following statement regarding the ten-mile rule: "Most Delegations agreed to a width of ten miles, provided a system were simultaneously adopted under which slight indentations would not be treated as bays" (p. 12). Several attempts were made to define the characteristic configuration of a bay The German whose waters should be regarded as national or interior waters. delegation, for example, proposed measuring the maximum depth of a bay in proportion to its breadth from headland to headland. The British delegation proposed taking into account the ratio between average depth and breadth by measuring the area. Subsequently the British and German delegations withheld their amendments and voted for the American proposal.

The American proposal avoids the definition of such words as "bay" and "estuary" in a geographical sense. It simply undertakes to determine when an indentation of the coast is sufficiently great to regard the waters within the indentation as national waters, which are to be separated from territorial waters by a straight line drawn across the entrance.

The border-line case between an open bay whose waters are territorial waters, and a closed bay whose waters are national or interior waters, is assumed to be a semicircle whose diameter does not exceed ten miles (see Fig. 4). Because of the frequent irregularity of bays, instead of reckoning the simple or the average depth in proportion to the width of the entrance, it is proposed to take the general shape into account.

Since bays frequently have minor indentations which should be ignored for practical purposes, and since in the case of estuaries it would be difficult to determine how far up the river to go in measuring the area, the American proposal is to use a method inside the indentations which is exactly similar to the drawing of the arcs of circles from all points along the coast (see Fig. 5). It is drawn, however, not with a radius of three miles but with a radius which is proportionate to the width of the entrance. A comparison is then made between the area enclosed by the envelope of the arcs of circles and the straight line across the entrance (the shaded areas in the diagrams), and the area of a semicircle whose diameter is proportionate to the width of the entrance. When the area of the special "envelope" inside the bay exceeds the area of the semicircle, the waters inside the straight line are national waters, and the three-mile limit is measured from the straight line. Otherwise the limit is measured from all points on the coast.

In addition to the distinct advantage of being equally applicable to estuaries and bays, this expedient has the further advantage of both exaggerating and simplifying the characteristic shape of the bay or estuary. It is much easier to take into consideration the area enclosed within the envelope of the arcs of circles than it would be to take account of the area of the entire bay or estuary following all the sinuosities of the coast.

The American proposal with reference to bays and estuaries was submitted to the Hague Conference in the following language:

* *

"Subject to the provisions of Article * with reference to bays and other bodies of water which have been under the jurisdiction of the Coastal State, in the case of a bay or estuary the coasts of which belong to a single State, or to two or more States which have agreed upon a division of the waters thereof, the determination of the status of the waters of the bay or

estuary, as interior waters or high sea, shall be made in the following

manner:

"(1) On a chart or map a straight line not to exceed ten nautical miles in length shall be drawn across the bay or estuary as follows: The line shall be drawn between two headlands or pronounced convexities of the coast which embrace the pronounced indentation or concavity comprising the bay or estuary if the distance between the two headlands does not exceed ten nautical miles; otherwise the line shall be drawn through the point nearest to the entrance at which the width does not exceed ten nautical miles;

"(2) The envelope of all arcs of circles having a radius equal to onefourth the length of the straight line across the bay or estuary shall then be drawn from all points on the coast of the mainland (at whatever line of sea level is adopted on the charts of the coastal State) but such arcs of circles shall not be drawn around islands in connection with the process which is next described;

"(3) If the area enclosed within the straight line and the envelope of the arcs of circles exceeds the area of a semi-circle whose diameter is equal to one-half the length of the straight line across the bay or estuary, the waters of the bay or estuary inside of the straight line shall be regarded, for the purposes of this convention, as interior waters; otherwise they shall not be so regarded.

"When the determination of the status of the waters of a bay or estuary has been made in the manner described above, the delimitation of the territorial waters shall be made as follows: (1) if the waters of the bay or estuary are found to be interior waters the straight line across the entrance or across the bay or estuary shall be regarded as the boundary between interior waters and territorial waters, and the three-mile belt of territorial waters shall be measured outward from that line in the same manner as if it were a portion of the coast; (2) otherwise the belt of territorial waters shall be measured outward from all points on the coast line; (3) in either case arcs of circles of three-mile radius shall be drawn around the coasts of islands (if there be any) in accordance with provisions for delimiting territorial waters around islands as prescribed in Article * *

The reason for using a radius which is a given fractional part of the breadth of the bay between headlands, or where it first narrows to ten miles, is that it takes full account of the shape of both small and relatively large bays.

It may be found that a radius equal to one-fourth the length of the straight line across the bay may prove to be too large a fractional part, in that it may generalize the shape of the bay too much and not take sufficient account of minor indentations. It might be modified, for example, by using a radius equal to one-fifth of the length of the straight line in describing the arcs of circles within the bay, and then comparing the area within the envelope of the arcs with a semicircle whose diameter is equal to three-fifths of the length of the straight line across the bay. It will be necessary to try out the method on many bays, of different types, in which different countries are interested, before the acceptability of the American proposal with reference to bays and estuaries, and the desirability of modifying it, will be known.

The reason for ignoring islands in the process suggested in the American amendment is that it is impossible to take account of islands in or near the mouth of a bay and at the same time consider the real shape of the bay itself. Islands of all sizes and shapes are found in and near bays and it is believed to be impracticable to treat some of them as if they were mainland when some islands, because of their small size and position, would necessarily be ignored. The existence of islands in bays is taken into account quite independently in the American proposal for the elimination of anomalous pockets of high sea.

ASSIMILATION OF OBJECTIONABLE POCKETS OF HIGH SEA

If the general rule of describing the arcs of circles is generally accepted, and if an impartial rule is adopted for the classification of bays and estuaries as interior waters or high sea, it might seem that there would be no actual need of further elaboration of the method of delimiting territorial waters.

Nevertheless, it will be found that when the arcs of circles of three-mile radius have been drawn from all points on the coast-line of both mainland and islands, and when it has been determined what indentations of the coast have the configuration of closed bays whose waters are interior or national waters,