was reduced to 10 feet. Buffalo Bay of Lake of the Woods, the "meridian line", and Northwest Angle Inlet were similarly mapped. Horizontal control for the maps was furnished by triangulation and traverse (p. 114) and vertical control by lines of levels, stadia traverse along the boundary, and vertical angles (see below). The methods used by the United States surveyors differed somewhat from those used by the Canadians, but where areas overlapped for short distances the results were surprisingly similar. The United States parties used the plane table exclusively with a field scale of 1: 45,000, while under Canadian technique the topography of mountainous regions was developed by phototopography supplemented by traverses of about the same character as that used by the United States parties. On the sections of low relief the Canadian parties used a specially devised instrument for combined level lines and stadia traverse. This was accompanied by notebook sketches and hand-level readings. Notes and sketches were plotted and adjusted on field sheets kept in camp, thus affording opportunity for correcting errors and for filling in any overlooked areas. VERTICAL CONTROL The vertical control for the topographic maps includes elevations of bench marks that had previously been established by other agencies on or near the boundary, lines of levels run from previously established bench marks to convenient points on the boundary, and, where the character of the country made it practicable, lines of levels along the boundary from monument to monument. At the time the surveys for the boundary maps were made and the boundary levels run, most of the bench marks available for initial elevations were in unadjusted circuits with elevations subject to corrections, to be later determined, to reduce them to true mean-sea-level datum. Since that time, the United States Coast and Geodetic Survey and the Geodetic Survey of Canada have each run many miles of first-order levels paralleling, and in places crossing, the boundary, often connected at common bench marks. Each of these organizations has completed the adjustment of its first-order net of levels, and elevations of bench marks of each organization are now known on slightly differing mean-sea-level datums. The bench marks of the boundary survey and the bench marks upon which they were based have been tied to these first-order bench marks in many places, making it possible to adjust the elevations used in the boundary mapping to mean-sea-level datum as determined by the Geodetic Survey of Canada. The elevations published in this report (appendix IV, p. 218) have been adjusted, except for an instance or two noted, to agree with the elevations of the Geodetic Survey of Canada as published in 1929 and 1930.3 The elevations on the boundary maps were determined, as has been noted, and the maps were printed several years before data for the final adjustment of the elevations were available. Therefore, the elevations shown on the maps differ slightly from those published in appendix IV of this report. 3 Precise Levelling in Manitoba (Publication No. 21), Precise Levelling in Saskatchewan (Publication No. 22), Precise Levelling in Alberta (Publication No. 23), and Precise Levelling in British Columbia (Publication No. 24), Geodetic Survey of Canada, 1929 and 1930. For the vertical control of the maps of the section of boundary from Georgia Strait to the eastern base of the Rocky Mountains, a system of level lines in combination with vertical angles and stadia surveys was used, described as follows: On Point Roberts, where no bench marks were available, sea level from local observations was used as the datum for the mapping. No bench marks were established in this locality. Across the coastal plain from Boundary Bay to the base of the Cascade Mountains, bench marks of the United States Geological Survey, established in 1905 along and near the boundary from Blaine, Washington, to Sumas, Washington, were used for the control of stadia surveys of the topography. Since the mapping was done, the United States Coast and Geodetic Survey and the Geodetic Survey of Canada have both run first-order level lines to Blaine, connecting with each other and with the United States Geological Survey bench marks. The Geodetic Survey of Canada has established first-order bench marks on the boundary at Sumas-Huntingdon. At the boundary crossing of the Pasayten River a bench mark was established by running a checked line of levels from a United States Geological Survey bench mark at Barron, Washington, through Windy Pass and down the Pasayten River to the boundary. From Sumas, Washington, through the high mountain region to the crossing of the Pasayten River, elevations were determined by vertical angles and stadia traverse based on the elevations of the bench marks at Sumas and at the Pasayten River. From the Pasayten River eastward, elevations were likewise determined by vertical angles and stadia traverse to the Similkameen River, where connections were again made with United States Geological Survey bench marks. From the Similkameen River eastward to the eastern crossing of the Kettle River at Laurier, Washington, sufficient vertical control was provided by United States Geological Survey bench marks which since have been tied to first-order level lines of the United States Coast and Geodetic Survey and of the Geodetic Survey of Canada. From the eastern crossing of Kettle River at Laurier to the Columbia River crossing, elevations were obtained by vertical angles and stadia surveys based on bench marks at both places. The bench mark at the boundary crossing of the Columbia River was established by a double-run line of levels, 43 miles in length, from a United States Geological Survey bench mark at Meyers Falls, Washington. This line has since been tied to the first-order lines of both the United States and of Canada. From the boundary crossing of the Columbia River eastward to the western crossing of the Kootenai River at Porthill, Idaho, elevations depend entirely on vertical angles and stadia surveys. However, since the boundary maps were made, a highway from Spokane, Washington, to Nelson, British Columbia, crossing the boundary about 21⁄2 miles east of Clark Fork (Pend-d'Oreille), has been built and a bench mark established at the boundary crossing by the United States Geological Survey. The Geodetic Survey of Canada has checked the elevation of this bench mark. Barron a mining camp near Slate Creek, no longer exists. At the west crossing of the Kootenai River at Porthill, Idaho, a bench mark was established by running a double line of levels from a Great Northern Railway elevation at Bonners Ferry, Idaho, through Porthill to a Canadian Pacific Railway elevation at Creston, British Columbia. Since the boundary maps were made, the Geodetic Survey of Canada has tied to the Porthill bench mark from the north with a first-order level line, and the United States Coast and Geodetic Survey has tied to it with a first-order level line from the south. From the west crossing of the Kootenai at Porthill to the east crossing of the Kootenai at Gateway, Montana, the elevations on the boundary maps depend on vertical angles and stadia surveys. At Gateway, bench marks were established by a double-run line of levels from Rexford, Montana, based on a Great Northern Railway elevation. This line has since been rerun by the United States Coast and Geodetic Survey as a part of their first-order levels. From the east crossing of the Kootenai at Gateway to the North Fork of the Flathead River, the elevations on the boundary maps depend on vertical angles and stadia surveys. Bench marks were established on the Flathead River by a double-run line of levels from a United States Geological Survey bench mark near Apgar, in the vicinity of Belton, Montana, on the Great Northern Railway datum. This line has since been reduced to first-order datum by a connection with first-order levels of the United States Coast and Geodetic Survey at Belton. Across the summit of the Rocky Mountains to the valley of the St. Mary River at their eastern base, vertical angles and stadia surveys were the means of determining elevations for the topographic mapping. In the St. Mary River valley near the boundary, the United States Geological Survey and the United States Reclamation Service had established a number of bench marks from lines of levels run in circuits from the Great Northern Railway in the vicinity of Fort Browning and Cutbank, Montana. These circuits have recently been reduced to first-order datum through connections with the first-order levels of the United States Coast and Geodetic Survey at Fort Browning and at Cutbank. This combination of levels, stadia surveys, and vertical angles sufficed for mapping, on a 100-foot contour interval, the rugged country from Georgia Strait to the eastern base of the Rocky Mountains. At this point the abrupt change in the character of the country demanded that a much smaller contour interval be used eastward to accurately portray the topography. This required much more accurate vertical control and necessitated the running of almost continuous lines of levels along the boundary. From the United States Geological Survey bench marks in the valley of the St. Mary River, a single line of wye levels was run continuously along the boundary eastward for 62 miles to a connection with United States Geological Survey and United States Reclamation Service bench marks near the crossing of the Great Northern Railway at Coutts-Sweetgrass. Permanent bench marks were established all along the boundary, either in the bases of or near the boundary monuments. This line has now been adjusted to close on a Geodetic Survey of Canada first-order bench mark at Coutts. From Coutts, Alberta, eastward for 207 miles to Monument 478, about 4 miles west of Frenchman Creek, wye levels were run along the boundary and the elevation of each monument determined, but no permanent bench marks were set. Later, the United States Geological Survey established about 25 bench marks along this section of the boundary and many of these were used to check the boundary levels. The descriptions and elevations of these United States Geological Survey bench marks are included in appendix IV. The elevation of Monument 486 was determined by a line of levels about 36 miles in length run from a United States Geological Survey bench mark designated "Rock Creek triangulation station", about 12 miles north of Hinsdale, Montana. The elevation of the bench mark in Monument 486 has since been determined by firstorder levels by both the United States Coast and Geodetic Survey and the Geodetic Survey of Canada. From Monument 486 westward to Monument 478 a double line of levels was run and a permanent bench mark established at each monument. The elevation of Monument 538 was determined by running a line of levels approximately 40 miles in length from a United States Geological Survey bench mark near Custer's ranch, about 20 miles south of Scobey, Montana. The line was continued westward along the boundary and closed on Monument 486. A permanent bench mark was established at each boundary monument. The elevation of Monument 615 was determined by a line of levels approximately 52 miles in length run from a United States Geological Survey bench mark at Ray, North Dakota. Levels were then run westward along the boundary from Monument 615 to a closure on the bench mark at Monument 538. A permanent bench mark was established at each monument. The elevation of the bench mark at Ray has since been determined by first-order levels of the United States Coast and Geodetic Survey. From Monument 615 eastward, a line of levels was run to Monument 630 at Portal, North Dakota, a distance of 22 miles. A permanent bench mark was established at each monument. Since the time the boundary levels were run, the elevation of the bench mark in Monument 630 has been redetermined by first-order levels of both the United States Coast and Geodetic Survey and the Geodetic Survey of Canada, and this first-order elevation has been used in adjusting the boundary levels to the westward as far as Monument 486, where the elevation was determined by first-order levels in like manner. From Monument 630 to Monument 693, the elevation of each monument was determined by a line of levels along the boundary, but no permanent bench marks were established. From Monument 693 to Monument 721, across the Turtle Mountains, the boundary maps were made in 1910 and 1911. On account of the rough character of the terrain, the level lines run in connection with the mapping were not run along the boundary itself but along roads paralleling the boundary and at some distance from it; and no permanent bench marks were established. In 1911 the Geodetic Survey of Canada established a first-order bench mark in the base of Monument 693, and in 1917 established a bench mark of the same order in the base of Monument 721. In 1918 the Boundary Commission, in connection with checking some triangulation, ran a line of levels along the boundary between the two monuments and determined the elevation of a point on the base of each intervening monument. From Monument 721 to Monument 832, levels were run continuously along the boundary and the elevation of the ground at each monument was determined. No permanent bench marks were established. Since the time these levels were run, the Geodetic Survey of Canada has established several first-order bench marks on or near this section of the boundary. From Monument 833 at Red River to Monument 911 at Lake of the Woods, a distance of approximately 90 miles, a continuous line of levels was run along the boundary and a permanent bench mark was established in the base of each monument. This section of the boundary crosses a comparatively flat terrain including many miles of the Great Roseau Swamp where accurate leveling is extremely difficult. The leveling instrument used on this work was a specially designed Bausch and Lomb dumpy level equipped with a micrometer screw under one end of the telescope to keep the level bubble centered, and with a mirror by which the observer watches the level bubble while reading the rod. The levels were run in short circuits, and made to close within a limit of 0.05 foot multiplied by the square root of the length of the circuit in miles. The boundary levels within this section were run in 1912. They were started from the Geodetic Survey of Canada first-order bench mark 3-C in the foundation wall of the post office at Emerson, Manitoba, and were closed on a United States Geological Survey bench mark in Warroad, Minnesota, about 7 miles south of the boundary. The following year, 1913, the Geodetic Survey of Canada ran a first-order level line from Emerson, Manitoba, along the Canadian Northern (now Canadian National) Railway through Warroad to Rainy River, Ontario, and established bench marks along the line. They determined the elevation of Monument 909 and of the United States Geological Survey bench mark in Warroad. The boundary levels have been adjusted to agree with the adjusted values of the elevations of these first-order bench marks. In 1929 and 1930 the Topographical Survey of Canada, in the course of special topographic work, redetermined the elevations of Monuments 853 to 903, inclusive, through a system of levels based on the Geodetic Survey of Canada first-order bench marks between Emerson and Sprague. With the exception of certain instances where the monuments have been disturbed either by settling or by heaving, due to the action of frost in the swampy districts, these redetermined elevations are in good agreement with those originally determined by the boundary levels. For the reason that a number of the monuments had been disturbed since the boundary levels were run, the elevations determined by the Topographical Survey of Canada for Monuments 853 to 903 have been adopted and published in the list of "Elevations along the 49th Parallel Boundary", appendix IV, page 240. The vertical control for the meridian line, Monument 913 to Monument 925, was included in the vertical control for the entire boundary survey of Lake of the |