Two hundred and fifty-three islands

A glacial basin that did not empty cleanly

Lake Winnipesaukee ↗ holds two hundred and fifty-three islands. That number comes from the official survey maintained by New Hampshire, and it does not shift with the season: in this context an island is land that stays above the waterline year-round, not a shoal that disappears at spring runoff. The precision matters because the lake sits in a glacial basin — a depression gouged and then overdeepened by ice, its floor uneven in ways that have nothing to do with drainage and everything to do with the weight and movement of a glacier that retreated roughly twelve thousand years ago. What the ice left behind was not a clean bowl but a complicated one, and the islands are the high points of a submerged landscape that was never smoothed out.

The basin runs roughly twenty-one miles by nine, with a surface area of seventy-two square miles, making Winnipesaukee the largest lake in New Hampshire and the third largest in New England. Its maximum depth is about 180 feet, but that figure misleads: large portions of the lake are far shallower, and the islands cluster where the bedrock ridges come closest to the surface. The same glacial action that carved deep trenches between the ridges left the intervening high ground barely submerged, which is why the southern and central portions of the lake — Meredith Bay among them — are crowded with islands while other stretches are open water with almost none.

What the glacier made, and how it froze

Glacial lake basins ↗ in New England tend to have irregular bathymetry for the same structural reason: the glacier did not erode uniformly. It followed the path of least resistance, which in this part of the White Mountains foothills meant exploiting joints and fault lines in the granite underneath. Where the rock was fractured, ice found purchase and deepened the channel; where it was massive and intact, ridges survived. The result at Winnipesaukee is a bottom that would look, if drained, like a mountain range at moderate scale — peaks, valleys, benches — with the two hundred and fifty-three islands as the summits still in air.

The shallows between islands freeze before the open water does. That is not folklore; it follows from basic physics. Shallow water loses heat to the air more quickly than deep water, which retains thermal mass longer into the autumn. The channels between the closer-set islands — barely a canoe-length across in some places — can carry a serviceable sheet of ice weeks before the main lake consolidates. Local ice fishermen have always known this, and the progression of freeze across the lake maps almost exactly onto the bathymetric chart: freeze the shallows first, then watch the ice work outward into depth. Full consolidation across the lake typically happens in January, though the open-water trenches between major ridges may resist longer.

This freeze pattern has one structural consequence that is easy to overlook: island communities, even small ones, lose their informal water access — a rowboat, a small outboard — before the ice is thick enough to walk on. There is a window of weeks during which neither passage works reliably, and island residents have always had to time their movements around it. Bear Island, which carries a permanent population and its own post office, is the clearest local example of an island organized around exactly this logistical reality. The seasonal gap between liquid water and safe ice is not an inconvenience; it is the structural fact that shapes how island life is provisioned and maintained.

The count, the names and what was left unnamed

Of the two hundred and fifty-three islands, a majority are unnamed and uninhabited. Many are too small to support a structure, or sit on rock with insufficient soil for a well or a septic system, or are simply too exposed to prevailing winds to be usable for part of the season. The named islands skew toward the larger and the historically occupied: Bear Island, Governor's Island, Rattlesnake Island, Cow Island, Long Island. The names themselves are a shallow record of European settlement — practical designations assigned in the eighteenth and nineteenth centuries, sometimes noting what the island was used for, sometimes simply noting what it looked like from the water.

The Abenaki had a different relationship to this lake. The name Winnipesaukee is itself derived from an Abenaki phrase, most often translated as "smile of the Great Spirit" or, more literally, as something closer to "good water at the outlet" — the second translation aligns better with the lake's geography, given how Winnipesaukee drains south through the Weirs and into the Merrimack watershed. What the Abenaki called the individual islands, if anything, is not systematically recorded in the documentary sources available through the New Hampshire Division of Historical Resources. The European survey counted, named and in many cases platted islands that had been navigated for centuries without a formal count.

The surveyed number — two hundred and fifty-three — has its own history. Earlier figures varied. Nineteenth-century gazetteer entries put the count lower, partly because survey methods were imprecise and partly because the definition of "island" was applied inconsistently. A rocky ledge that clears the waterline by a foot in August might be charted as an island by one surveyor and ignored by another. The current figure reflects a standardized definition applied with modern survey equipment, and it is the number that appears on USGS mapping of the lake.

The archipelago is not distributed evenly. Meredith Bay clusters islands in its southern portion, with the shallows between them — the mill bay's approach from open water — creating a navigational puzzle for anyone unfamiliar with the lake. The steamer Mount Washington, which has worked the lake since the nineteenth century, follows a charted course that threads between the island groups rather than crossing the shallow ridges directly. That course has changed very little since the first surveys, because the bottom has changed very little. Twelve thousand years of post-glacial time have deposited a thin layer of sediment over the bedrock ridges, but not enough to reroute a working vessel. The glacier set the geometry; the survey counted what it made; the islands remain where the ice left them.