Winter, Monhegan Island

Step 1 · The picture’s history

A finished picture by April; an undated afternoon

Two fishermen in a dory, fish houses beside the harbour, snow across the foreground: the Met’s 1922 catalogue describes a late-afternoon view, with “cold yellowish lights and deep blue shadows”. Its phrase “the effect is icy.” describes the painting’s effect. It is not a report that the sea was frozen or a reading of the air temperature.

The Met identifies the distant islands as Manana, Smutty Nose and Little Smutty Nose. It records Kent’s arrival on Monhegan in 1905 and includes this canvas in his April 1907 New York exhibition. New Britain’s museum essay dates the Clausen Galleries opening to 1 April; the period Met catalogue transcribes the signature and year.

We therefore examine 1 January–31 March 1907, the calendar months before the exhibition in the signed year. This is seasonal context, not a documented execution interval. The sources examined do not establish whether this canvas was painted outside, worked up in a studio, or informed by observations from an earlier winter. A signed year cannot date every experience behind a picture.

Step 2 · The regional record

Snow that remained on the ground

At Portland, Maine, measurable snow lay on the ground at observation on 86 of 90 days in January–March 1907. The other entries are 1 trace and 3 zeros without a trace flag. The record concerns the snow already lying there, not how often snow fell.

The largest reported depth in this interval is 61.0 cm, on 27 February–1 March 1907. Using NOAA’s station coordinates and the Monhegan light station as a location reference gives a straight-line separation of 76.1 km. Portland is a mainland station with a different exposure. These values do not measure the island’s snow, the harbour’s water or the light Kent saw.

Step 3 · Test the comparison

Persistent in this archive; local conditions unknown

The 86-day count ranks 3 of 36 complete January–March windows in Portland’s 1893–1940 snow-depth archive, with the most persistent first. This is a comparison within the published record. Continuity of the observing site and snow exposure has not been established, so the rank is not a homogenized measure of climatic rarity.

Every comparison uses the same calendar months with 29 February omitted. A trace is not measurable depth. A missing reading, a failed quality check, or a value marked as presumed zero excludes that year; missing snow is never counted as no snow. No modern normal is attached: this station’s record ends in 1940, and no airport series is spliced onto it.

Step 4 · What the records can say

The snow has context; the light remains Kent’s

The regional record makes persistent snow a reasonable context for a canvas completed by early April 1907. It cannot establish that Kent painted a particular snowfall, or that the snow cover on Monhegan matched Portland. The catalogue’s description puts his treatment of cold in the relationship between warm light and blue shadow. We have no dated source view from which to say that he added snow, changed the sky or faithfully recorded one afternoon.

Kent later remembered intense cold on Monhegan. In Wilderness, the entry for 16 November 1918, he recalled water freezing near his stove during his first island winter. That is the artist’s retrospective testimony, not a weather observation dated to this painting. Neither the memory nor the Portland rank supplies the missing painted day.