General Land Office survey plats for the Beaver Island archipelago

Summary

The project holds 11 scanned General Land Office survey plats covering the Beaver Island archipelago. They are public-domain federal survey records served by the Michigan Department of Natural Resources, distinct from the BLM patent and field-note data that remains blocked behind the General Land Office portal.

The plats make the early federal survey visible as a map rather than only as a modern GIS layer. The images show the Public Land Survey System section grid, water boundaries, survey annotations, and historical labels such as BIG BEAVER I. and GARDEN ISLAND. The provenance records describe the underlying originals as approved circa 1849. An exact survey or approval date should still be taken from the title block of the individual plat being cited.

What the plats establish

The collection establishes that the archipelago was represented in the federal survey framework during the period when the Strangite settlement was active. It also provides a direct cartographic comparison point between the earlier GLO labels, Farmer’s 1853 map, and the 1854–55 Lake Survey chart. That comparison is especially valuable for names such as Big Beaver, Garden Island, and interior lakes whose later histories are otherwise reconstructed from maps and narrative sources.

The plat scans do not establish who owned each parcel. A survey plat lays out the public-land framework and records survey information. The separately held patent layer now identifies recorded patentees, issuance dates, and legal descriptions; tract books and county deeds are still needed for entry history and later chains of title. See BLM General Land Office patent records for the Beaver Island area.

Joinable geography

The held PLSS layers provide the geometry needed to connect the plat township, range, and section labels to modern GIS. The state collection goes further to quarter-quarter polygons. The patent import now uses that join: 170 legal-location rows have exact quarter-quarter matches and 495 remain section-only because the source describes lots or fractional subdivisions, or the finer geometry is not held. The project does not infer ownership from geometry alone.

The KB exposes that join layer as 125 PLSS section records under kb/articles/documents/plss-section-*. Those records are derived from the machine-readable federal section geometry and link back here for comparison with the scanned plat sheets. They should not be read as 125 independent ownership claims or as substitutes for the plat images.

The individual scans are also indexed as 11 plat-specific records under kb/articles/documents/glo-plat-*. Those records preserve what is legible on each sheet without treating the whole collection as one undifferentiated date: the original township plats are described as circa 1849, while the second sheet of T39N R09W records a 1942 survey and 1943 acceptance for Pismire Island, and the second sheet of T39N R10W records a 1954 supplemental plat for Section 11 based on the 1849 plat.

The next join layer is now indexed under kb/articles/documents/plss-quarter-quarter-*: 1,153 aliquot-part records at the resolution most useful for future patent and deed matching. Current DNR parcel records are indexed under kb/articles/documents/dnr-parcel-*; they preserve legal descriptions and agency classifications while keeping private ownership unresolved. Current OSM named and historic features are indexed under kb/articles/documents/osm-named-feature-* and are map evidence, not title evidence.

How to traverse the plat network

The 11 individual records are designed as a navigable evidence layer rather than as isolated PDF notes. Each record connects outward in four different ways:

• Survey geometry: its township/range identifier links to the matching federal PLSS section records, including the section-11 record used by the 1954 T39N R10W supplement.
• Environmental reconstruction: modern PLSS geometry is used to find the circa-1800 land-cover units within the township extent. Those units derive from GLO surveyors’ section-line notes, but the spatial join does not claim that every polygon is visible on the scanned sheet.
• Names and places: GNIS feature points and their exact name-record rows provide a current gazetteer comparison. A GNIS link is not proof that the current name appeared on the circa-1849 plat.
• Map chronology: neighboring held plats and the 1855 Lake Survey chart provide continuity across sheet boundaries and an independent later map comparison.

This separation makes the graph useful for questions such as “which survey sections and later names fall within the T39N R10W sheet?” while preserving the different evidentiary status of an image observation, a modern geometry join, a reconstructed land-cover classification, and a current gazetteer record.

Open questions

• Which exact survey and approval dates remain legible in the original title blocks?
• Does the plat for T39N R10W label Font Lake, or does that name first appear on the 1853 Farmer map?
• Which named features on the plats correspond to current GNIS features, and which are only historical labels?
• Can the still-missing tract books and surveyors’ field notes be obtained through the replacement BLM system or an archival copy? The patent layer is now held and linked through BLM General Land Office patent records for the Beaver Island area.

See also

• BLM General Land Office patent records for the Beaver Island area
• North End of Lake Michigan Including the Beaver Island Group (1855 Lake Survey chart)
• Current named-feature and property join records
• Historical propertyholder and ownership research queue
• Land survey and settlement geography
• Strang's biblical toponyms
• Lumbering