line than a natural one. The idea of a bank is of water limited by dry land with some elements of permanency about it. Even during the brief period when the line is uncovered, the idea of it is suggestive far more of mud and aquatic growths than of dry land and forest growths.

"To my mind, the natural, simple and obvious idea of the bank of a lake, in this climate, is presented only by the line of mean high water. Here we would first find permanent dry ground every day of an average year. Here an observer, during every annual round of ordinary seasons, would see the water advance to his very feet and then recede, as if some power had drawn the line and said to the waters, hitherto shalt thou come but no further.' Here the struggle between forest growths and aquatic vegetation begins to change the landscape. Here lines of drift, the flotsam and jetsam of the waves naturally suggest the limits of the 'bed of the lake.'

"Without doubt, then, I conclude that mean-high-water mark determines the shore of the lake; and it now remains to designate that level, and how it shall be found.

"Several surveys of the proposed Nicaraguan Canal route, beside that of Commander Lull above quoted, have been made within the last fifty years. Each found a certain mean high level of the lake, and it might seem a simple solution to take an average of them all. But, as each adopted its own bench-mark on the ocean, and ran its own line of levels to the lake, I have no means of bringing their figures to a common standard. It seems best, therefore, to adopt the figures of that one which is at once the latest and most thorough, which has enjoyed the benefit of all of the investigations of all of its predecessors, and whose bench-marks on the lake are known and can be referred to. That is the survey, still in progress, under the direction of the United States Canal Commission. Its results have not yet been made public, but, by the courtesy of

The highest level of the lake in 1898 was 106.7, last of November. The elevation of our bench mark on inshore end of boiler at San Carlos is 109.37. . . .'

"A complete copy of this letter will be handed you, and also blue-prints of the maps made by the Commission of the lower end of the lake, which may facilitate your work.'

Ás this Commission is the highest existing authority, I adopt its finding, and announce my award as follows:

"The shore line of Lake Nicaragua, at the level of 106.0 feet, by the bench marks of the United States Nicaragua Canal Commission, shall be taken as the bank of said lake referred to in the treaty of 1858."

The location of the line around the southern margin of the lake was the most difficult part of the whole survey. The country here is a vast morass, densely covered with tropical vegetation, even the sluggish streams being mostly choked with aquatic plants. The high water line was defined as 106 feet, while the level of the lake at the time of the survey was about 102. It became necessary, therefore, to determine an elevation four feet higher, which was usually several miles from the water's edge.

By means of levelling, it was ascertained that the swamps had a mean slope toward the lakes of about one foot per mile, and this was made the basis of most of the agreements. A few points on streams were located, the area of Nicaragua's two-mile zone was calculated, and

cent form, the centre would be, not among them, but within the convex space which they partially surround.

Other methods must therefore be devised for finding the mid-position of irregular and restricted areas, and many might be suggested, more or less applicable to different figures. But it will be sufficient here to indicate only the method which I have adopted as best suited to the figure in hand, possessing, as this does, something of a curved or crescent shape.

"I have supposed a vessel to enter the Bay from the ocean, at a point midway between its headlands, and to sail a course as nearly as possible equidistant between the opposite shores, on the right and left, until it has penetrated to the remotest point of the Bay.

This course, being carefully plotted upon the map, although curved, may be taken as the long axis of the Bay.

"At right angles to it, at different points, I have drawn straight lines reaching across the Bay from shore to shore, and, by use of a planimeter, I have determined the position of such a line which will exactly divide the whole area of the Bay into equal parts. This line may be taken as the corresponding short axis of the Bay, and its intersection with the long axis will be the centre of the Bay.

"When at that point, a line drawn across the bow of the supposed vessel, perpendicular to her course, would have one-half of the waters of the Bay in front of it and one-half behind it.

"Having carefully located the point in this manner, I have determined from the scale of the map, its distance from the summit point of the small island in the Bay, whose latitude and longitude are given upon the map as follows:

Latitude, II° 03' 10"

Longitude, 85° 43′ 38′′

"It proves to be 37 seconds to the northward and 14 seconds to the eastward of this point.

"I therefore fix the position of the centre of Salinas Bay to be:

HE route for the Nicaragua Canal as projected by the present Isthmian Canal Commission is shown on the accompanying map. It generally follows the course of the San Juan River for one hundred miles from the Caribbean Sea to Lake Nicaragua about one hundred and five feet above it, then it traverses the lake for a distance of seventy miles to the mouth of the Rio Las Lajas, and after following the valley of that stream for a short distance, crosses

the continental divide, forty-four feet above the lake, and descends the valley of the Rio Grande to Brito, seventeen miles from Lake Nicaragua.

The canal as proposed will have a mean depth at low water of thirty-five feet and a bottom width of one hundred and fifty feet. This width is for the straight sections; on curves with a radius of less than 12,000 feet the width is increased at the rate of one foot for each two hundred feet; thus a curve with a radius of 6,000 feet

of Lake Nicaragua, secured by means of an immense masonry dam which will extend the level of the lake fifty miles down the San Juan. This stage consists of improved river and lake channels. Third, a period of descent from the lake level to the Pacific through the continental divide. This stage, seventeen miles, must also be excavated.

The salient engineering problems connected with the Nicaragua Canal project as outlined by the Commission are as follows:

First. The construction of harbors at the termini of the canal.

good harbor at Greytown, the eastern terminus of the canal, with thirty feet of water at the anchorage and about the same depth in the entrance. The entrance to this harbor where it then existed, has been obliterated and the harbor itself is now a lagoon almost entirely enclosed, of restricted area, with only about half the depth of water in it that formerly existed. Vessels for Greytown are now compelled to anchor in the offing and discharge their cargoes on lighters which are taken into the lagoon across a bar having a depth of less than six feet of water. As the prevailing trade winds are

strong and blow almost directly on this part of the canal, the construction of a harbor at this terminus becomes necessary in the early stages of the work as well as for use after the canal is completed.

It is proposed to construct, by excavation, a harbor of sufficient area to accommodate vessels using the canal. The entrance would have a minimum depth of thirty-five feet and a bottom width of five hundred feet, guarded by two jetties springing from the shore line near Harbour Head. These jetties are to be built of loose stone to a height of six feet above mean high tide, the hearting to be composed of small and the outer portion of large stone, not easily moved by the waves. It is not expected, however, that the construction of the jetties will alone form the entrance. Dredging will also be necessary and its maintenance may re quire an extension of the jetties or dredging or both.

The western terminus of the canal will be near Brito. Here, as at Greytown, Here, as at Greytown, there is no harbor, and an artificial one must be constructed. The same general engineering principles will guide in its construction. The width and depth of the

entrance will be the same. The sand movement on the western coast, however, is slight as compared with that in the vicinity of Greytown. The prevailing winds on this side are off-shore, and destructive storms seldom visit this part of the coast. The cost of maintenance of the harbor on the west side will therefore be less than that of the one on the east side at Greytown.

For a part of the distance between Greytown and the Florida Lagoon the canal line passes over swampy sections, where the material is too soft to support the embankments necessary to keep out the floods of the San Juan, and to maintain the canal level itself. Protecting embankments are therefore to be constructed over these sections. These embankments are to be located as far as practicable on the firm land composing the neighboring hills. In places, however, they cross ground which is soft to a considerable depth. Waterways are provided on the embankment lines to dispose of flood water in the protected areas.

At two places near the Boca San Carlos dam site heavy cutting is encountered, the maximum depth for short distances being