Second in a series on water boundaries.

Water boundary principles are easy to state in the abstract but difficult to apply in practice. The usual way to analyze how a particular legal rule applies to a situation is to consider the reason for the rule and determine whether that purpose is furthered by applying it in the instant case. For example, a rule states that monuments control courses in a description. When you think about the "whys" behind that rule, you come up with concepts of visible utility being preferable to abstract measurements, principles of certainty, relative permanence, and the like. But if in a particular instance the use of a monument does not further those aims, then it probably should be rejected.

With water boundaries the reasons behind the rules are complex, elusive and often arbitrary. Why, for instance, does a call for a stream in a metes and bounds description mean that the location of the stream controls over a series of meander courses following the stream? Usually, the courts say that the stream is a natural monument and leave it at that. However, the same court may apply the rule that a lateral shifting of the stream by accretion moves the boundary, but a movement by avulsive action leaves the boundary at the meanders. Obviously there is more at play here than the certainty of visible monuments being preferable to abstract measurements. The law may view a stream as having a relatively permanent existence, but it will never consider the stream's location to be permanent. Consequently it throws in a balancing between the importance of maintaining stream access for a riparian owner with the societal value of allowing people to adjust the use of their property to follow slowly moving natural boundaries but not rapidly moving ones. This leaves you with more questions, such as "How fast is an avulsion and when does it start?" I once read a court report where the judges blithely stated-in reference to a western river that had slowly shifted its banks until it was blocked by a log jam, whereupon it jumped out of its channel and formed a new one-that the boundaries along the river had moved until the "original" avulsion, when they became stationary. Of course a river that jumps its banks once will probably do it again and again (and had before), so I was left wondering what possible means a surveyor on the ground would have to determine which move was the "original" avulsion?

The following cases address questions of defining the mouths of rivers, bays and inlets, and (in the first case) of determining the ownership of islands accreting within riverbeds. In both instances the courts began, in earlier cases over the same boundaries, with broad generalizations that followed "settled" boundary rules, and thus needed little justification for their conclusions. Unfortunately the parties still had plenty of room-and time-to argue the details. Equally unfortunately, by the time the court got to the hard questions it tended to take the easy road, merely stating that it felt that one party had the better argument without giving much of an explanation as to why. Nevertheless, it is useful to see how the Supreme Court addresses these issues.

The Cases Georgia v. South Carolina, 497 U.S. 376 (1989).

This case concerns the boundary between the two states along the lower reaches of, and at the mouth of, the Savannah River. The court states that the colony of Georgia was created by letters patent in 1732, describing the boundary with the existing colony of South Carolina as "the most northern part of a stream or river there, commonly called the Savannah." 497 U.S. at 380. The Treaty of Beaufort (1787) later added clarifying statements: "The most northern branch or stream of the river Savannah from the sea or mouth of such stream to the fork or confluence of the rivers now called Tugoloo and Keowee, and from thence the most northern branch or stream of the said river Tugoloo till it intersects the northern boundary line of South Carolina... reserving all the islands in the said rivers Savannah and Tugoloo to Georgia;...shall forever hereafter form the separation limit and boundary between the States of South Carolina and Georgia." The court further described the treaty as granting navigation rights in the north part of the river to both states, with such rights in the southern part going to Georgia, and also that the treaty did not state whether the boundary was the middle of the northern branch of the river, the South Carolina shore, or whether the bed was held jointly. 497 U.S. at 381.

In 1876 the Supreme Court held that the treaty did not prevent the federal government from regulating commerce in the river. In 1922 the court issued several statements interpreting the Treaty:

"The Court held, among other things, that (1) where there is no island in the Savannah River, the boundary is midway between the banks when the water is at ordinary stage, (2) where an island is present, the boundary is midway between the island bank and the South Carolina shore, with the water at ordinary stage, (3) where a navigable or nonnavigable river is the boundary between the two States, and the navigable channel is not involved, then, in the absence of contrary agreement, each State takes to the middle of the stream, and (4) the location of the boundary under the Treaty was unaffected by the thalweg doctrine because of the Treaty's provision that each State shall have equal rights of navigation." 497 U.S. at 383-84.

This case concerned jurisdiction over several islands, the location and size of the mouth of the river, and the lateral seaward boundary between the two states.

The court first discussed the Barnwell Islands, a group of four islands claimed by South Carolina by prescription. In 1955 the Fifth Circuit Court of Appeals had adjudicated a dispute, originally filed in Georgia, over a federal government action to condemn an easement on one of the islands. The private owner of the island moved to dismiss the complaint, alleging that proper jurisdiction was in South Carolina. The Fifth Circuit ruled that the specific reservation to Georgia of islands in the river settled any question that Georgia had jurisdiction. Nevertheless, the Supreme Court in this action awarded the Barnwell Islands to South Carolina:

"We need not here repeat in detail the extensive record evidence and the tax and conveyancing documents relied upon [by South Carolina]. It suffices to say that the entire area in the late 18th and early 19th centuries was low marshy ground. The islands were separated from Georgia by the wide and deep waters of the Savannah River, but were separated from South Carolina only by streams so shallow that they were described as sometimes dry." 497 U.S. at 392.

The shallow streams between the islands and South Carolina had accreted in the 20th century, largely due to navigation maintenance activities in the main river channel by the U.S. Army Corps of Engineers. Now the disputed territory comprised at least 450 acres of ground just downstream from the city of Savannah. South Carolina had issued the private land grants on the island and collected taxes. It had seized certain property on the islands for nonpayment of taxes and had policed and patrolled the islands, enforcing its game laws. Georgia noted that there had never been much activity on the islands, except for that of the Corps and for a Confederate gun battery during the Civil War. As far as private activities are concerned, Georgia alleged that there had been a little rice growing and illegal whisky production, but little else taking place on the islands.

Prescription of territory by one state against another requires acquiescence of the state being dispossessed. Georgia essentially argued that with so little private activities taking place on the islands there was nothing for it to acquiesce to, especially after the Fifth Circuit decree in 1955. The Supreme Court saw otherwise, holding that the fact that South Carlina had always collected taxes on the islands should have put Georgia on notice that their inaction would result in their losing jurisdiction.

The next issue, concerning islands that emerged in the Savannah River after the treaty, was more difficult from a water boundary point of view. Remember that in 1922 the court had ruled that the state boundary ran down the middle of the river, and down the middle of channels running north of any islands in the river. The court began with a synopsis of relevant issues from its 1922 decision:

"Two principles established by the 1922 decision are pertinent here. First, although it is by no means self-evident on the face of the Treaty that the 'northern branch or stream' refers to the 'stream' that each island-however small and however close to the northern shore-creates between itself and the shore to the north of it, that was the construction of the Treaty agreed upon by the parties in 1922 and adopted by this Court. Apparently it was thought that a contrary rule, whereby the 'northern branch or stream' referred only to a 'branch or stream' that made a major departure from the main body of the river, would create an unmanageable boundary, because the Treaty's additional reservation of the islands to Georgia would create pockets of Georgia territory within South Carolina wherever islands existed on the South Carolina side of the 'northern branch or stream' defined in this larger sense. Second, under the principle that each island in the river created a new 'northern branch or stream,' each island was not only reserved to Georgia under the reservation clause of Article I, but also formed a point of reference, by which the boundary would be drawn." 497 U.S. at 395-96.

Georgia, in this case, argued that the above quoted principles applied equally to islands that emerged in the river subsequent to the treaty:

"Georgia's solution, whereby each emerging island not only is newly 'reserv[ed] ... to Georgia' but also creates a new 'northern branch or stream' by which the boundary between the States must be drawn, would create a regime of continually shifting jurisdiction. Even the smallest emerging island, no matter how near the South Carolina shore, would cause the entire boundary between the states to shift northward, depriving South Carolina not only of the land that constitutes the island but also any riverbed between the island and the center line that previously formed the boundary." 497 U.S. at 396.

The court disagreed with Georgia's view. It felt that the treaty was intended to fix the boundary, rather than create a system whereby the emergence of an island would cause radical shifts in the line. Such a construction also "comports with principles of simplicity and finality." 497 U.S. at 397.

"We recognize, of course, that the normal rules relating to accretion and erosion may cause the boundary line between the States to shift over time, so that the line will not necessarily be fixed as of any particular point. But it is one thing to say that the parties meant that gradual shifts in the path of the river would shift the boundary gradually, to the extent of the accretion; this rule is consistent with settled expectations and with the parties' interest in maintaining their riparian rights...It is quite another thing to infer that the parties meant that each new island, however formed, would alter the boundary line to a degree that could be dramatically out of proportion to the physical change brought about by the formation of the island itself." 497 U.S. at 397.

The court concluded that newly emerged islands belonged to the state owning the particular portion of the riverbed where the island arose.

The next issue turned on the location of the river mouth. Georgia argued that it owned Oyster Bed Island, which the court stated is one of the most downstream islands in the Savannah River. It is worthwhile to note that there are numerous islands in and near the mouth of the Savannah. Some of these are bounded on one side by the Savannah and on other sides by small streams or sloughs which could arguably be part of the river or not. Georgia had conceded that Turtle and Jones Island, which lie on the north side of the north channel of the Savannah, were part of South Carolina. Georgia contested, however, the recommendation of the Special Master (who took the evidence in this case) that downstream of Turtle Island the boundary should run southeast to the main navigation channel of the river and then to the mouth. Georgia argued that in the 1870s the main channel ran north of Oyster Bed Island and was shifted south by U.S. Army Corps of Engineers' activities to keep the southerly channel open. The court held that Oyster Bed Island was part of South Carolina, basing its decision not on the shifting channel but on the proper location of the mouth of the river.

Tybee Island forms the south side of the mouth of the river. The problem lies in defining the north side. Georgia argued that the north side of the mouth should be defined by Daufuskie Island and Hilton Head Island. These lie several miles to the north of Tybee and arguably are islands because of being cut off from the mainland by sloughs, rather than by the Savannah itself. In short, the problem here lay in deciding where the river ended and the ocean began.

"Usually, there are two opposing 'headlands' marking and constituting the mouth of a river.... This is the 'headland-to-headland' principle used in defining the limits of bays and rivers.... It is not always that simple, however. Sometimes the mouth of a river is difficult to delineate.... Because of the absence of a reasonably close headland to the north, Georgia is driven to argue that the boundary at the mouth of the Savannah River must be the geographical middle between Tybee and the closest points of land in South Carolina, that is, Daufuskie Island, lying north and northeastward of Turtle Island, and Hilton Head Island, almost six miles north of Tybee.

"We conclude that this is not a realistic determination of the Savannah River's mouth, and we agree with the Special Master in rejecting the argument.

"The difficulty lies in the fact that Tybee Island, the most seaward point of land on the southern side of the river, has no counterpart of high land on the northern side. The geographical feature taking the place of the customarily present opposing headland is, instead, a shoal, long recognized as confining the river. It is true, of course, that the Corps of Engineers affected the flow by its training wall and hydraulic fill. But the shoal which directed that flow has been recognized for many years. Furthermore, Hilton Head Island and Daufuskie Island are so far distant that it is impossible to say that they even touch the Savannah River." 497 U.S. at 399-400.

Georgia won the next issue, which concerned the way the boundary in the river should shift northward when the river encountered an island. The court, in its 1922 decision, had held that the boundary was the middle of the river except where there was an island, in which case the boundary would be the midway between the island and the South Carolina shore. The court, in this case, was asked to say how to construct the line where the boundary moved at each "end" of the island. The Special Master suggested that the line run midway between the island and the South Carolina shore until the island ended, and then that a line be constructed at a right angle to either the island to bank centerline or the bank to bank centerline, whichever was shortest. The court disagreed.

The major consideration for constructing a line down the "middle" of a river is that every point on the midline be equidistant from the nearest points on the opposite shores. 497 U.S. at 401. Extending this principle, the court felt that the best solution was to follow Georgia's recommendation, that "with an island's presence, the boundary is to be marked by the use of a point which is tri-equidistant from the South Carolina shore, the island shore, and the Georgia shore. The boundary then would pass through this point and otherwise be equidistant from the South Carolina shore and the Georgia shore, or island, as the case may be." 497 U.S. at 401.

Georgia also won the next contest, concerning land added to Denwill Plantation (in South Carolina) and to Jones Island. Before the 1880s, the navigation channel that ran around the north side of Elba Island, which lies just upstream of Denwill Plantation, was very wide. The U.S. Army Corps of Engineers improved the channel by narrowing it through the construction of a training wall north of Elba Island. This narrowed and deepened the channel, and also deposited material against Denwill Plantation and Jones Island, which became dry land extending past the original state boundary in the middle of the river as it existed in the 1880s. Most of the construction work by the Corps took place from 1891 to 1895, but the deposition in the river took over 40 years. 497 U.S. at 402.

Which state owned the new land (where it extended south of the old boundary) depended on whether the change was avulsive or amounted to an accretion to the South Carolina shore. The court decided the change was avulsive, and awarded the land to Georgia:

"General rules concerning the formation of riparian land are well developed and are simply expressed and well accepted. When the bed is changed by the natural and gradual processes known as erosion and accretion, the boundary follows the varying course of the stream. But if the stream leaves its old bed and forms a new one by the process known as avulsion, the result works no change of boundary.... Sometimes, the problem is to distinguish between the two.

"Here we have a situation where interference in the river's flow was not caused by either of the adjoining states, but by the United States Army Corps of Engineers. It is generally held, of course, that one cannot extend one's own property into the water by landfilling or purposefully causing accretion....

"We conclude, not without some difficulty, that Georgia has the better of the argument as to these two areas. It is true, of course, that avulsive action ordinarily calls to mind something somewhat sudden or, at least, of short duration, whereas accretion has as its essence the gradual deposit of material over a period by action of water flow. This is so even though it may have been caused partly or wholly by placed obstructions.

"Some of the changes here were caused gradually by the deposit of sediment by river waters. Others were caused by the deposit of fill through the use of a hydraulic-pipeline dredge employed by the Corps pursuant to the paramount right of the United States Government to improve navigation... The rapidity of some aspects of the dredging and other processes led the Special Master to conclude that the changes in the Savannah River were primarily avulsive in nature. Although the question is close, on balance, we think this particular record as to this particular river supports the recommendation made by the master." 497 U.S. at 403-04.

The final major issue concerned the lateral seaward boundary. The court had already decided that the mouth of the Savannah was defined by Tybee Island on the south and a "long recognized shoal" (which lies about a mile from Tybee) on the north. The Special Master then continued the midline of the river past the mouth until it intersected a straight line running from the most northerly point on Tybee to the most southerly point on Hilton Head Island. He then ran the line seaward in a direction perpendicular to the line between the points on the two islands. 497 U.S. at 406.

Georgia and South Carolina both took exception to the Master's line. Each wanted the overall course of the entire coastline of its state considered, and the line run perpendicular to that. According to the court, the general run of the Georgia coastline was about North 20º East, while the general run of the South Carolina coastline was about North 47º East. Perpendiculars to these two courses would overlap, and each state preferred the solution based on its own coastline. The court adopted the Master's line. "We conclude that it gives equitable balance and recognition to the so-called equidistant principle and to the inland boundary between the two States, and does so with the least possible offense to any claimed parallel between offshore territory and the coast itself." 497 U.S. at 408.

United States v. California, 382 U.S. 448 (1965)

I do not plan to detail the arguments of this long running dispute over the ownership of offshore lands abutting California; I merely want to quote the Court's definition of "inland waters." A significant baseline in this case was the "coast line," which the court defined as the line of mean lower low water and the line marking the seaward limit of inland waters. Thus the location of the line across the mouth of an inland water is significant as it forms part of the coastline:

"4. The inland waters referred to in paragraph 2(b) hereof include: (a) Any river or stream flowing directly into the sea, landward of a straight line across its mouth; (b) Any port, landward of its outermost permanent harbor works and a straight line across its entrance; (c) Any 'historic bay,' as that term is used in paragraph 6 of Article 7 of the Convention, defined essentially as a bay over which the United States has traditionally asserted and maintained dominion with the acquiescence of foreign nations; (d) Any other bay (defined as a well-marked coastal indentation having such penetration, in proportion to the width of its entrance, as to contain landlocked waters, and having an area, including islands within the bay, at least as great as the area of a semicircle whose diameter equals the length of the closing line across the entrance of the bay, or the sum of such closing lines if the bay has more than one entrance), landward of a straight line across its entrance or, if the entrance is more than 24 geographical miles wide, landward of a straight line not over 24 geographical miles long, drawn within the bay so as to enclose the greatest possible amount of water. An estuary of a river is treated in the same way as a bay.

"5. In drawing a closing line across the entrance of any body of inland water having pronounced headlands, the line shall be drawn between the points where the plane of mean lower low water meets the outermost extension of the headlands. Where there is no pronounced headland, the line shall be drawn to the point where the line of mean lower low water on the shore is intersected by the bisector of the angle formed where a line projecting the general trend of the line of mean lower low water along the open coast meets a line projecting the general trend of the line of mean lower low water along the tributary waterway." 382 U.S. at 450.


Obviously some of the questions addressed above are specific to the situation. Whether an island in the Savannah River was gained by a neighboring state by prescription, or whether land deposited as a result of navigational improvements by the Corps of Engineers amount to an avulsion, are more questions that can only be answered in reference to individual case facts. One does, however, need to remain cognizant that such questions will come up and be prepared to meet them.

How to draw the lines across the mouths of rivers and bays, as well as around islands, is of much more universal interest, and is seldom addressed in texts. This question comes up most often where two bodies of water, treated differently with respect to boundaries, come together. For example, the boundaries along a non-navigable stream might be at the thread, while the boundaries along a navigable river might end at the ordinary high-water line. The difficult question arises in locating the boundaries where a non-navigable stream empties into a navigable river or lake. Some jurisdictions might extend the plane of the navigable river's ordinary high-water line upstream into the smaller stream; others might draw a straight line across the mouth of the stream. The solutions presented in these cases, which are about as clear and as precise as any you are likely to find in court reports, should give you food for thought.

To study this topic further, I recommend reading chapters six and seven of Water Boundaries by George M. Cole (John Wiley and Sons Inc., New York, N.Y., 1997). That book does a good job of illustrating and explaining the application of many of the principles discussed in the cases quoted above, as well as several variants on the theme.