the marks had sufficient strength to resist this destructive action. The top of the base was made hexagonal in shape and the concrete was sloped down at an angle of 45° to solid rock on all sides. On this base the range mark was built, using only the upper 4-foot section of the portable form. Small steel I-beams were used for reinforcement throughout, and wherever the rock was smooth, the structure was anchored to the foundation with steel tie rods.

Two other special marks were built of the same design as the others, but differing in size. One, range mark 8, on the Battery at Eastport, Me., incloses the old rubble-masonry range mark as a core. This mark as reconstructed is 12 feet high, and each side of the triangular base measures 12 feet. The other, range mark 28, at Lubec, Me., although located on the best available site for a rear range for the south end of boundary course 7-8, is hidden from view by buildings. It is about 41⁄2 feet high. A white stripe 3 feet wide was painted on the side of the building facing range mark 27 of this pair to serve as an auxiliary range mark for this course.

TOPOGRAPHY

The character of the country along the boundary from the source of the St. Croix River to the Atlantic Ocean varies from the rolling farming country along the St. Croix River below Woodland, Me., to the wooded swamps of the upper part of the St. Croix River Valley. The valley of the St. Croix River between Woodland and Vanceboro, Me., is heavily wooded and rolling, with few knolls which rise more than 100 feet above the river. Above

Vanceboro the boundary traverses the Chiputneticook Lakes, whose outstanding features are their long narrow bays and numerous islands. Above North Lake the boundary follows Monument Brook, a narrow winding stream which drains a swamp.

When operations were begun on Passamaquoddy Bay and the St. Croix River in 1909, maps which were accurate and modern, in the sense intended in the treaty of 1908, had been made of part of this region, but they were based upon surveys made for purposes other than those of boundary demarcation. They included marine charts of Passamaquoddy Bay published by the British Admiralty and by the United States Coast and Geodetic Survey; a chart of the St. Croix River from its mouth to the head of navigation at Calais, Me., and St. Stephen, New Brunswick,

and Geodetic Survey; and unpublished field maps, based upon surveys made by the United States Coast and Geodetic Survey, upon which were shown the shore line and soundings made to locate the channel from Calais to the mouth of Monument Brook. The published

maps of the United States Coast and Geodetic Survey were contoured on the United States side of the boundary, from West Quoddy Head to Calais, and on the Canadian side from Warwig Creek to St. Stephen but above Calais and St. Stephen no contouring had been done.

The commissioners decided that the information furnished by these maps should be used in preparing the "accurate modern charts" required by the treaty, and that they would do such additional topographic work as might be necessary to make the maps of this section of the boundary conform to the standards fixed for the maps of the other sections. Accordingly during the course of the surveys made for the determination of the geographic positions of the monuments and turning points of the boundary, topographic maps were made of a strip of country along each side of the boundary from the source of Monument Brook to the international bridge at Calais, Me., and St. Stephen, New Brunswick, and on the Canadian side of the boundary from Warwig Creek to Grand Manan Channel. The topography of the remaining territory, along Passamaquoddy Bay and the St. Croix River, was taken from the charts of the United States Coast and Geodetic Survey after a careful revision had been made of the semipermanent topographic features.

The field maps were made on various scales from 1: 5,000 to 1: 20,000, depending upon the scale upon which the final maps were to be published, and upon the detail necessary to show the relation of the boundary line and reference monuments to the topographic features. A 10-foot contour interval was used on the field maps north of Baring, Me.; a 20-foot interval south of that point.

In the field the plane-table-and-stadia method was used almost exclusively for mapping the topography, the area on each shore being mapped to a minimum width of one-fourth mile. The horizontal control for the topographic survey was furnished by the boundary triangulation and traverse, which determined the geographic positions of the reference monuments. The vertical control was furnished by a system of spirit levels run in closed circuits from precise-level bench marks, but for immediate control the elevation of the water in the lake or river was often used as a starting elevation for the plane-table traverses.

In open country the various natural and artificial topographic features were mapped by plane table, telescopic alidade, and stadia. In the more densely wooded areas this method was supplemented by the use of open-sight alidades, paraffined

linen tapes, and aneroid barometers, in accordance with the method developed by the United States Geological Survey for filling in the topography of wooded areas around which control points had been established. White celluloid field sheets were used, and on these a projection was first laid down and then the positions of triangulation and traverse stations and monuments were plotted. The shore line as shown on the United States Coast and Geodetic Survey field maps was then transferred to the field sheets and was later checked in the field and corrected if necessary. Closed plane-table-and-stadia traverses were run between control stations, advantage being taken in the wooded regions of old logging roads and river drivers' trails. The contours, roads, streams, and other topographic features were located by numerous stadia readings taken from the traverse stations.

At the time the topographic maps were being made in the field hydrographic data were also obtained to determine the position of the main channel or thalweg of the St. Croix River in the vicinity of the numerous islands. These waterways were thoroughly sounded and the position of each sounding determined and plotted on the plane-table sheet by alidade and stadia. These hydrographic sheets were then used by the commissioners as the basis of information in laying down the boundary line.

H

Special type of dumpy level used on St. Croix River boundary

LEVELING

The vertical control for the topographic maps consists of lines of second-order 5 levels run in closed circuits and based upon the elevations above mean sea-level datum of first-order level bench marks of the United States Coast and Geodetic Survey and precise-level bench marks of the Geodetic Survey of Canada.

The elevation of a bench mark and of the water surface of Grand Lake at Butterfield Landing was determined by running a double line of levels from United States Coast and Geodetic Survey first-order level bench mark J, (elevation 391.577 feet) at Danforth, Me. From Butterfield Landing a single line of levels was run to Orient, Me., where a bench mark was established. The elevation of this point was checked by running a side line to Grand Lake at Peters Landing. From Orient lines were run in closed circuits to a bench mark at The Thoroughfare and to the source of Monument Brook, where a bench mark was established in the base of Initial Monument.

The water level of North Lake was determined from the bench mark at The Thoroughfare; and, from the lake, levels for topographic work were carried up Monument Brook to the bench mark at Initial Monument.

Bench marks were established at the extreme southern end of Grand Lake and at Spruce Mountain Cove of Spednik Lake by running a line of levels in closed circuits along the highway from United States Coast and Geodetic Survey first-order level bench mark N, (elevation 440.022 feet) at Forest Station, Me. Water-level gauges were established near these last two bench marks, and from them the daily variation in the elevation of the lakes was determined, so that the water surface could be used by the topographers as a reference plane.

5 The requirement for closure of circuits for levels of this class at that time was that the discrepancy in closure in feet should not exceed 0.05 √M, in which M is the length of the circuit in miles. This was the criterion for each small circuit of the line.

As a check on the elevation of the water surface of Spednik Lake, a gauge was established near Vanceboro at the southern end of the lake. The elevation of the zero mark of this gauge was determined by running a closed circuit of levels from United States Coast and Geodetic Survey first-order-level bench mark V, (elevation 391.594 feet) at Vanceboro.

A line of second-order levels was run in circuits from this United States Coast and Geodetic Survey bench mark at Vanceboro down the St. Croix River, to Gleason Point. Thence the main line of levels was run along the roads on the Canadian side to Milltown, and, crossing the river, it was terminated at Calais, Me., at preciselevel bench mark 1-B (elevation, 53.578 feet) of the Geodetic Survey of Canada. From this main line of levels, various branch lines were run in closed circuits and permanent or temporary bench marks were established at points suitable for the vertical control of the maps.

Elevations for topographic use on the Canadian islands of Passamaquoddy Bay were based on a tidal bench mark of the United States Coast and Geodetic Survey stamped No. 3 (elevation 37.35 feet) at Eastport, Me. The elevation of a temporary bench mark at Welshpool, New Brunswick, was determined by readings of the water surface at high tide made simultaneously at Eastport and at Welshpool. From this temporary bench mark spirit levels were then run along the principal highways, and other temporary bench marks were established for the topographers. The elevation of a temporary bench mark on Deer Island was also determined in a similar manner from the bench mark at Eastport.

The leveling instrument used along the St. Croix River was a specially designed Bausch & Lomb dumpy level which had been used on the bogs along the MinnesotaManitoba boundary. This level was equipped with a mirror, which enabled the observer to watch the level bubble while reading the rod and by means of a micrometer screw one end of the telescope could be slightly raised or lowered to keep it level while the reading was being made. A target rod was used with a vernier reading to thousandths of a foot. The target, which was designed by the engineer to the United States section of the commission, comprised, in addition to the regular form of target, a small scale on which the stadia distance was read by the upper stadia wire after the target was set at the correct elevation. To prevent the possible error of setting the target on one of the stadia wires instead of on the center wire, the lower stadia wire was removed from the instrument, thereby making it impossible to read the stadia distance if the target was improperly set. The target is shown in the illustration on page 91.

A list of the permanent bench marks established, together with their elevations and descriptions, is given in Appendix III (p. 166).

FIELD TRANSPORTATION

The transportation of equipment, supplies, and monumenting materials and the transfer of men to and from their work each day was not a particularly difficult problem for the parties which surveyed this portion of the international boundary, for Passamaquoddy Bay and the greater part of the St. Croix River are navigable either for motor boats or for rowboats and canoes.