Charting and survey

Summary

The archipelago was charted twice over, by two federal programs that rarely overlapped except at the water’s edge. The US Lake Survey and its successor agencies mapped the lake bottom and the shoreline as a hazard to shipping, beginning with a trigonometrical survey run under Capt. J. N. Macomb in 1854 and 1855 and continuing, sheet by renumbered sheet, into NOAA’s current chart 14911. The US Geological Survey mapped the land as terrain, beginning only in 1955-56 — nearly a century later — because no federal topographic survey of the islands existed before then. Commercial map publishers such as John Farmer supplied the only general-purpose maps before the Lake Survey arrived, and a private compiler, George Scott, supplied the sailing directions that turned chart soundings into compass courses a pilot could steer.

What the held charts establish is narrower than a complete survey history. The project holds 186 catalog records for charts intersecting the archipelago, but has read only a handful — the 1855, 1900, 1907 and 1911 sheets — at full resolution; the rest remain catalog-level stubs recording title, chart number, year and scale only. Within that limit, three findings are traceable across the held record: the dedicated Beaver Island Group sheet increased in scale from 1:120,000 to 1:80,000 between 1907 and 1911 and held that finer scale for at least a century afterward; the general small-scale chart of the whole lake ran continuously from 1898 to 2024 under a changing chart number and only began carrying the label “Mercator Projection” in its title with the 2002 edition; and the 1956 topographic quadrangle drew its hydrography directly from 1955 Lake Survey charts, tying the land-mapping program to the lake-charting one at the one point they touched.

The 1854-55 US Lake Survey

The earliest instrumental survey of the archipelago held in this archive is the US Lake Survey’s “North End of Lake Michigan Including the Beaver Island Group,” compiled from a trigonometrical survey run under Capt. J. N. Macomb, Topographical Engineers, in 1854 and 1855, at a scale of 1:120,000 and engraved by W. H. Dougal for the Topographical Bureau of the War Department. The chart itself, in the held 1861 duplicate impression, names the field officers by assignment: Macomb ran the general triangulation and offshore hydrography; Lieut. W. F. Raynolds surveyed Hog Island, Garden Island and the north end of Beaver Island in 1854; assistants W. H. Hearding, J. H. Foster and J. L. Beghem covered the remainder of Beaver Island and the north coast west of Point Patterson in 1855; and assistants G. W. Lamson, H. Gillman and sub-assistant J. B. Williams surveyed High, Gull, Trout, Whiskey and Squaw Islands in 1855. This is the earliest systematic American survey record held for High Island, Garden Island and the island now called Ojibwa Island.

The chart is not only a shoreline outline. Its “Light Houses” table records tower height and visible range for the four lights then established (no Squaw Island light yet appears, matching the survey credits), and its “Sailing Directions” box gives compass courses and statute-mile distances between Beaver Harbor, Waugoshance Light, Manitou Passage, Green Bay and Chicago — a direct predecessor of the tabulated routes George Scott would publish 47 years later. It also shows, though it does not label, an inland road-and-clearing pattern running south from Saint James that later sheets would identify as King’s Highway. The held duplicate carries a handwritten 1861 presentation inscription to the “Geographical & Statistical Soc’y of N.Y.,” signed in print by “Geo. G. Meade, Capt. Top’l Engineers, Sup’t Survey N. and N.W. Lakes” — the same Meade who would command the Army of the Potomac at Gettysburg two years later, then Superintendent of the US Survey of the Northern and Northwestern Lakes, based in Detroit.

Revising the Beaver Island sheet, 1897-2024

The chart-catalog metadata held for the archipelago (186 records intersecting the area of interest) lets the same sheet be tracked across chart numbers and decades in a way no single chart image can show by itself. The dedicated large-scale coverage of the group runs, in sequence: LS10 (1855, reissued 1897, 1:120,000); LS0 (1900, 1:120,000); LS143 (1903, 1:120,000); LSMa (1907, 1:120,000); then LS72 (1911), the first edition at the finer scale of 1:80,000; LS79 (1917, 1923, 1928, 1929, 1931, 1935, all held at 1:80,000); and, after a renumbering into NOAA’s modern national chart series, chart 14911, “Waugoshance Point to Seul Choix Point, including Beaver Island Group,” held in editions from 1975 through 2023, still at 1:80,000. The scale increase between the 1907 and 1911 editions is the one clear change in resolution the catalog documents for this sheet; the finer scale then held for at least 112 years. No marginal note on either sheet explains the reason for the 1911 change, and it is recorded here as an observed fact, not an explained one.

The 1900 edition of the earlier LS0 sheet carries its own evidence of a different kind of revision: a bottom-margin correction notice stating that aids to navigation were corrected from information received to 1 February 1900. This is a running-corrections practice distinct from a full resurvey — updating lights, buoys and other aids on an existing plate between full triangulation campaigns — and it is the only such notice read directly from a held sheet in this archive so far. The 1907 and 1911 sheets, read at full resolution for the 1907 and 1911 KB records, also carry inland detail beyond soundings: both show the same broken road pattern south of Saint James that the 1855 chart shows unlabeled, and the 1911 sheet is the first to print the label “Kings Highway” on it. The 1907 sheet’s own authorities panel attributes the underlying topography chiefly to the 1854-55 Lake Survey work, with 1902-05 resurvey work limited to hydrography and shoal soundings — so a cleaner 1907 or 1911 rendering of an inland feature is corroboration of the 1855 topography, not proof that the feature was independently resurveyed on those later dates.

The wider net: general and regional charts

Alongside the dedicated Beaver Island Group sheet, the Lake Survey and its successors published smaller-scale charts covering the whole lake or a wider regional stretch of coast, for a different navigational purpose — plotting a passage rather than entering a harbor. A “General Chart of Lake Michigan,” at a constant scale of 1:500,000, runs under chart numbers LS83 (from 1898), LS131, LSM and LS7 through 1970, then LS7M, then, after the same modern renumbering, chart 14900/14901 from 1975 to 2024. The catalog title for this lineage first carries the phrase “Mercator Projection” with the chart-14901 edition dated 2002; earlier editions in the same lineage are titled simply “General Chart of Lake Michigan” or “Lake Michigan.” Whether the earlier editions used a different projection, or the same one under a different title, is not established by the catalog metadata alone and is recorded below as an open question.

A third, intermediate-scale sheet covered the archipelago’s southern approach: chart LS78, “Lake Michigan — Frankfort to Charlevoix, Mich., Including Manitou and Fox Islands,” held in editions from 1912 to 1938 at 1:120,000. Its 1912 edition was read directly and confirmed to show the southern portion of Beaver Island alongside the Manitou and Fox Islands, even though neither island group is named in its title. Three scales therefore covered the same water at the same time for different purposes: roughly 1:500,000 for a lake-length passage, 1:120,000 for a regional approach, and 1:80,000 for entering the group itself. The single broadest-scale sheet held in the archipelago catalog is an untitled 1895 “Lake Michigan” chart at 1:666,947 — a catalog stub only, not read beyond its title and date.

Scott’s sailing directions, 1902

A chart records position; it does not by itself tell a pilot what course to steer or when a light should come into view. Scott’s New Coast Pilot for the Lakes, compiled by George Scott, “Licensed Pilot of all the Lakes,” from “personal notes and observations of over forty years’ experience,” supplies that text. The sixth edition, printed in Detroit by the Free Press Printing Company in 1902, gives Beaver Island group content on printed pages 139-141, 169, 262-265, 273, 275-276 and 310 — corresponding to PDF pages 190-192, 220, 313-316, 324, 326-327 and 361 in the held scan, which carries a roughly 51-page front-matter offset from its printed pagination.

The relevant pages give exact light characteristics and tower heights for the Beaver Island group’s stations, compass courses “corrected for magnetic variation” between them, a magnetic declination reading of “Beaver Island Light 0°.2 E.” for 1900, the warning that Gull Island Reef “bears S.S.E. ¾ E., 4 miles, from the south point of Gull island,” and the Life-Saving Service roster naming Owen Gallagher as keeper of the Beaver Island station. Compared against the 1855 Lake Survey chart 47 years earlier, the Coast Pilot’s data shows that at least two lights — Beaver Island Harbor Light and the head-of-island light — had been re-equipped in the interval, and that the Squaw Island light, absent from the 1855 chart’s own light table, existed by 1902.

From lake to land: US Geological Survey topography

No federal topographic map of the archipelago exists before 1955-56; the Lake Survey and its successors mapped the water and the immediate shoreline, but not the island interiors as terrain. The 1956 USGS Beaver Island quadrangle, at 1:62,500, closes that gap. Its marginal notes state the topography was compiled photogrammetrically from aerial photographs taken in 1954, field checked in 1956, with hydrography compiled from US Lake Survey charts 704 and 706 (1955) — the one documented point where the two federal mapping programs, lake and land, drew directly on each other’s data rather than working in parallel.

The 2023 US Topo quadrangle pair covering Beaver Island North and South, at 1:24,000, is a different kind of product: USGS’s own metadata describes US Topo as a GIS-derived repackaging of existing National Map and other government layers — orthoimagery, roads, contours, hydrography and names — rather than new photogrammetric compilation or field checking. The nominal scale is finer than 1956’s, but the two are not directly comparable production methods: one is a single continuously field-checked sheet from a specific aerial-survey campaign, the other a layered digital product assembled from several already-existing sources at a stated publication date.

Shoreline and bathymetric surveys since

Two later federal survey products extend the charting record without replacing it. NOAA’s National Geodetic Survey ran Shoreline Mapping Program project CM-9303F for Beaver and Garden Islands, producing a compiled vector shoreline whose line features carry a source date of 10 July 1993; the project’s own atlas record for this data states plainly that it is a dated source view, not a measurement of shoreline change, a current shoreline, or a legal boundary. Separately, NOAA’s National Centers for Environmental Information hold a lakewide integrated bathymetric-topographic grid and contour shapefile for Lake Michigan — the depth-sounding successor to the soundings drawn by hand on the 1854-55 chart, though the held record states no island-specific survey date for it.

Read the document

393 pages, rendered from the held scan.

Jump to a page — 393 in all

12345678910111213141516171819202122232425262728293031323337414549535761656973778185899397101105109113117121125129133137141145149153157161165169173177181185189193197201205209213217221225229233237241245249253257261265269273277281285289293297301305309313317321325329333337341345349353357361365369373377378379380381382383384385386387388389390391392393

135 of 393 pages linked here — every page to 32, then every 4th, then the last 16. With JavaScript, every page.

The page

Rendered from the held scan.

The page

Rendered from the held scan.

Added 4 August 2026. Recent changes