David P. Buckson, Atty. Gen., and Ruth M. Ferrell, Deputy Atty. Gen., for plaintiff.
James L. Latchum, of Potter Anderson & Corroon, Wilmington, for defendant.
DUFFY, Chancellor: 
This is an action for a declaratory judgment by the State of Delaware on the relationship of the Attorney General against The Pennsylvania Railroad Company ('The Pennsylvania') which counterclaimed for judgment in its favor. The controversy is concerned with title and related rights to certain 'foreshore,' which here is a strip of soil between high and low watermarks on the west side of the Delaware River.
The Court determined in prior opinions that (1) The Pennsylvania is a riparian owner and holds title to mean low watermark; (2) The Pennsylvania's right to fill its foreshore is not limited by any public right to navigate or fish in the waters of the foreshore; and (3) the State has not exercised such police power as it may have to preclude The Pennsylvania from filling and reclaiming the foreshore. State ex rel. Buckson v. Pennsylvania Railroad Company,228 A.2d 587 (Del.1967); 237 A.2d 579 (Del.1967).
This is the decision after final hearing, which involed two principal issues: (1) location of mean low watermark ('MLW') on the site and (2) whether The Pennsylvania has filled in the River beyond mean low watermark. Locating MLW means fixing the boundary line between the respective
properties of the parties. The State owns the bed of the River up to MLW, the Railroad's riparian title runs down to it. Where is the line that separates public from private property?
The disputed foreshore adjoins fast land of The Pennsylvania between Edgemoor and Bellevue. The terrain is a mudflat (flat, sloping land made of mud). It is washed twice daily by tidal cycles. Because of the nature of the terrain, its boundaries are not located by reference to any physical landmark on the site. And location is further complicated because the MLW is a 'shifting' line--it moves.
MLW is located at the intersection of two planes: the horizontal surface of the River and the vertical surface of the shore ('contour') when the surface of the River is at the elevation of mean low water as determined by observations for an epoch of 23 years beginning in 1942.  The intersection is established by an infinite series of points at which the surface of the River at the elevation of mean low water is in contact with the contour of the shore. The elevation of mean low water is determined by averaging all low tide readings taken during the 23-year period. The tidal datum so determined is thus a mathematical constant which does not change.  Hence the horizontal component or horizontal plane of MLW does not change. The vertical plane is inconstant and thus is the variable factor in determining the point of intersection.
The contour of the shore in question changes with change in the shoreline resulting from erosion or accretion of soil. As a consequence, MLW for a given date may be mapped at that time but, once change takes place, MLW cannot be identified by its former position. A MLW that existed in the past can be ascertained only by reference to a survey or map of that MLW made before the contour changed.
The MLW at the foreshore in question changed during the period between 1942 and 1967. And maps and surveys are the only guides to where it was at any given time. Those facts are fundamental in this lawsuit.
Changes in MLW are illustrated by a Composition Plan (put in evidence by the State) showing changes of mean low water line since 1942; it is attached to this opinion as Appendix 1.  As shown thereon, in 1942 MLW was at its most easterly (riverward) position. Its location at that time is determined by a Corps of Engineers survey made in 1942, as to which the parties are in agreement. In 1954 MLW was entirely west of 1942 MLW. By 1967 MLW had moved again; with respect to the 1954 position, it moved east in one area, west in other areas, and remained almost coextensive in still other areas. 
The State contends that the Court should declare 1954 MLW as the boundary line between the properties because that was the boundary at the time The Pennsylvania began filling the foreshore, and the eastward movement of MLW between 1954 and 1967 was not 'natural and gradual' (and hence should be disregarded under the 'moving boundary rule').  The State argues that its evidence as to the location of 1954 MLW is accurate and reliable.
The Pennsylvania contends that the 1954 MLW is irrelevant because the moving boundary rule does not apply. Alternatively, it contends that in any event the State's evidence as to the location of that line is too inaccurate for boundary purposes. The Pennsylvania then asks the Court to adopt the 1967 MLW as the boundary, because of its relevance and because that line is the only one which can be accurately located.
These contentions permit narrowing of the precise land area in dispute. Examination of the Composition Map shows that the disputed foreshore (Edgemoor to Bellevue) can be divided into five separate areas, each one of which is bounded on the east and west by the 1954 and 1967 MLW's, and on the north and south, by the points where the 1954 and 1967 lines intersect.
In the areas I have marked on Appendix 1 as Numbers 1, 3 and 5, 1954 MLW is East of 1967 MLW. Since The Pennsylvania has not filled east of 1967 MLW, it has not filled beyond the line which the State contends is the boundary, i.e., 1954 MLW. Moreover, as to those three areas, The Pennsylvania, by relying on 1967 MLW, concedes that such land belongs to the State. Hence, areas numbered 1, 3 and 5 are not in dispute.
In the area I have numbered 2, 1967 MLW is shown as lying east of 1954 MLW. However, the two lines (1954 and 1967) are so close together (1/8 of an inch apart on a map scaled at 1 inch equals 200 feet) that I regard them as being coextensive.
This leaves one area, which I have numbered 4, as the critical foreshore in dispute. It comprises about 7.7 acres and is set out separately on the Composition Plan. If 1954 MLW is the boundary, that area belongs to the State (and it follows that The Pennsylvania has filled on the State's land). If 1967 MLW is the boundary, that area belongs to the Railroad (and no trespass has been committed).
In support of its contentions the State makes various arguments to the effect that the natural water line is the boundary between the properties, the law of shifting boundaries is applicable, artificial filling does not change the legal boundary between the properties, and so on. But in the view I take of the case, these arguments are without legal relevancy. I say this because fundamental to the State's case is that, factually, the data shown on and interpolated and extrapolated from the 1954 survey by the Corps of Engineers is both accurate and reliable in locating the low water mark in question. That must be established before considering whether acceptance of 1954 MLW gives a fair and just result, as the State contends.
It is undisputed that all of the Railroad's filling is West of MLW as shown on the 1942 survey by the Corps of Engineers. But certain of the filling is, as I have already noted as to the 7.7 acres, East of MLW as described in the State's evidence based on the 1954 survey. The State relies upon that line, 1954 MLW, for judgment in its favor, and I regard it as the State's burden of proof to establish accurately for present purposes the location of that line by a preponderance of the evidence. Compare Reed v. Short, Sup. 1946, 5 Terry 103,57 A.2d 90. The State affirms that as the boundary line--it must locate it, reliably.
In my judgment the State has failed to meet its burden. I say this for the following reasons:
(1) 1954 MLW no longer exists and it is reasonable to infer from the evidence that it has not existed for many years. There is no physical evidence to identify where it was. 
(2) The best evidence as to the location of 1954 MLW is derived from the 1954 survey by the Corps of Engineers. I attach as Appendix 2 a copy of the pertinent part of that survey showing the area in dispute.
(3) That survey by the Corps of Engineers was not made for boundary purposes. It was made for dredging purposes. While that is not necessarily fatal, it does demonstrate the present limitations of the survey.
(4) The low water mark is shown on the survey by the dotted lines. The depths shown are taken from soundings made approximately 1,000 feet apart. By the scale on the survey, 1 inch equals 800 feet.  In locating MLW on the 1954 survey interpolation is possible in some locations but extrapolation is required in others because certain soundings are not shown thereon. A distance of 50 feet plots to 1/16 of an inch and the zero line is obviously 'freehand' drawn between the sounding data. And at other points extrapolation is required because there is no data above the zero line.
(5) A further adjustment must be made in order to use the 1954 survey. The datum thereon as to 0.0 elevation is 2.9 feet below the sea level datum of 1929. The Corps of Engineers for their purposes use 2.9 feet below sea level as the datum for their work on the Delaware River in this area. But in this case all parties are agreed that the datum (the horizontal plane of the River) is 2.20 feet below mean sea level. On the 1954 survey, therefore, mean low water corresponds to 0.7 foot elevation. MLW would have to be determined from the 1954 survey by plotting points representing 0.7 feet above the 'zero' level and then drawing a contour through these. And, as I have already noted, some of the sounding columns do not include elevations above the zero level and the addition of 0.7 feet would require plotting of the contour--extrapolation--with a guide line on only one side, i.e., without plus values. On the 800-to-1 scale, 0.7 feet is represented, according to my calculation, by less than nine ten-thousandths of an inch (.0009). The inherent difficulties in relying on such minute plotting on a survey not made for boundary purposes are self-evident.
(6) All of the witnesses testified as to some inaccuracy in using the 1954 survey for the purposes of determining MLW. They disagreed only as to the degree of error involved. According to one witness, the error would be approximately 20 feet; another testified that it would be as much as 100 feet. But none of them explained how he reached his conclusions as to the degree of inaccuracy. Therefore the Court has no evidentiary basis for concluding that a one-hundred foot error is less probable than a twenty foot error. And it would be sheer speculation on my part to assume that such errors would 'average out.'
(7) The scope of such possible error is illustrated by applying the evidence to the 7.7 acres in dispute; that area has a north and south width of approximately 1700 feet. By my calculations, a twenty foot error (20 1700 ) would involve between 4/5 and 5/6 of an acre. If the error is 50 feet, it would amount to some two acres, and if it is 100 feet, some four acres out of the total are involved.
I conclude that demarcation with the precision which the Court is called upon to make for boundary purposes cannot be fairly made from the 1954 survey by the Corps of Engineers. In short, the State has failed to establish by a preponderance of the evidence the location of MLW in 1954.
The question remains: Which MLW should be adopted as the boundary?
Under the evidence the choice is between the 1942 survey by the Corps of Engineers and the 1967 surveys made for the parties. The methods and procedures used in the 1942 survey are almost identical to those used in preparing the 1954 survey. The only significant difference between the two surveys is the absence in the 1954 survey of several soundings above the 'zero' line, thus requiring extrapolation to plot the relevant tidal datum 0.7 foot above 'zero'. In the 1942 survey no extrapolation is necessary, although interpolation would be required. But the same problems of scaling remain, and so do most of the others. There is no compelling evidence that the 1942 survey would be significantly more accurate (or less inaccurate) than the 1954 survey. For these reasons the 1942 survey is not acceptable for boundary purposes.
This leaves the 1967 surveys. The Court believes the proper and practical solution is to use 1967 MLW as evidenced by these surveys. They were made for boundary purposes. Their accuracy is not disputed. And the record is bare of any acceptable alternative. I therefore hold that mean low watermark along the entire site, which is the boundary line in question, is that located by the 1967 surveys. Each side submitted a different survey, but I understand the parties to agree that there is no significant difference between them. If there ...