Nephosthe weather behind the paintings

Every study, hung by year

84 studies from 1611 to 1954. Each is a page with its sources, its verdicts and what the record cannot say.

American painters

Also in American collections: Caillebotte, Pissarro and Monet at the Art Institute of Chicago, and most of Constable’s dated skies at the Yale Center for British Art.

Martin Johnson Heade, Newburyport Meadows: haystacks and light across a marsh.

Martin Johnson Heade · c. 1876–81

Newburyport Meadows

At least 5 local rain dates also record clear daylight during the hay seasons of 1879–80. Herald witnesses connect passing showers and wet weather with the harvest; no painted day is assigned.

Seasonal contextSky unknown

The Metropolitan Museum of Art, 1985.117

Women working beside the cottage doorways, with children and a cat on the steps.

Winslow Homer · c. 1881–1882

Sparrow Hall

Nephos reads a warm, sunny day. Durham records 38 days reaching 18°C with at least six sunshine hours in January–October 1882; 28 fell in July and August. Warm sunshine was a recurring summer possibility, but the canvas’s day and weather at the cottage remain unknown.

Warm sunshine testedPainted day unknown

National Gallery of Art, 2025.89.10

U.S. Weather Bureau daily map for 14 May 1892, north-eastern states. The painting itself is not reproduced.

Oliver Hazard Perry · 1892

View of Stonington, May 14, 1892, at 5am

Inscribed on the front in Perry’s hand. The Weather Bureau’s 8 a.m. map puts Block Island, 35 km away, at 44 °F under a clear sky; sunrise at Stonington was 4:29 standard time.

“5am” partially agreesPlace unknown

Florence Griswold Museum, 2017.1.3

Winslow Homer, After the Hurricane, Bahamas: a figure among wreckage on a beach under a dark sky.

Winslow Homer · 1899

After the Hurricane, Bahamas

NOAA’s hurricane database records no tropical storm or hurricane in the Atlantic between 4 November 1898 and 26 June 1899, the winter the museum gives for the painting.

Hurricane unknown

Art Institute of Chicago, 1933.1235

1611

Adam van Breen, Skating on the Frozen Amstel River: skaters before a stepped-gable house and a pink windmill on the frozen river.

Adam van Breen

Skating on the Frozen Amstel River

Dated “16011” on the gable, read by NGA as 1611; the winter shown is not dated

KNMI’s Low Countries reconstruction classes 1609–10, 1610–11 and 1611–12 as “Cold”, one step below “Severe”: a run of chilly winters, as an NGA article says. Fabricius logged a month of frost in East Frisia from late December 1610. The ice on the Amstel and the painted day remain unknown.

Chilly winters around 1611Painted day unknown

National Gallery of Art, Washington, 2010.20.1

 The painting at NGA

1802

c. 1808–1811

1816

John Constable, Weymouth Bay: a dark stormy sky over Bowleaze Cove.

John Constable

Weymouth Bay

Honeymoon at Osmington, about mid-October to the end of November 1816; the V&A sketch and the National Gallery canvas

A weather diary kept at Exeter, transcribed in full, shows Constable’s six weeks at Osmington running from a mild October into one of the coldest Novembers on record, with two stormy spells and fair, breezy days between hard frosts. Both pictures’ skies fit that autumn; whether the larger was painted then or recalled later, the weather cannot say.

The storm on the spot partially agrees1816 or 1819 unknown

Victoria and Albert Museum and National Gallery, London, 330-1888 and NG2652

 The painting at the National Gallery

1821

“September 3d. Noon. very sultry. with large drops of Rain falling on my palate light air from S.W.”

His note. The image is at the museum.

John Constable

Study of sky and trees

3 September 1821, date disputed, Noon

On the conditional 1821 date, Howard records SW and ‘Cloudy and fine at intervals.’ The wind direction partly agrees, but the sources do not establish the rain falling on Constable’s palette, the sultriness or the lightness of the wind.

wind partially agrees

Victoria and Albert Museum, 151-1888

1821–22

1821–1822

1822

1826

1830

1833

1836–37

1844

1851

1854

1856

1857

1859

1862

1866

1870

Gustave Courbet, La Falaise d’Étretat après l’orage: the cliff and arch above a calm sea, boats on the shingle.

Gustave Courbet

La Falaise d’Étretat après l’orage

Dated ’70; from Courbet’s stay at Étretat, August and September 1869

A month into Courbet’s stay, on the night of 11–12 September 1869, a Channel storm struck Le Havre, Fécamp and Dieppe, the most violent at Le Havre since 1832. The Observatoire’s bulletin logged a “tempête” at Le Havre that morning; nothing ties the canvas to that one storm.

A storm mid-stay partially agreesDay unknown

Musée d’Orsay, MNR 561

 The painting at the Musée d’Orsay

c. 1869/1872

About 1871

1872

Thomas Moran, The Grand Canyon of the Yellowstone: the Lower Falls in a sunlit yellow canyon.

Thomas Moran

The Grand Canyon of the Yellowstone

Completed in 1872 from sketches made at the Lower Falls, evening of 27 July to morning of 31 July 1871

The Hayden survey’s own weather log shows the week Moran sketched the falls as fair high-plateau summer: frost and 42° at dawn beside the canyon, afternoons near 70°, the sky about half clear by day, one storm on the horizon. The thermometer moved up-river the day he arrived.

The bright sky partially agreesDay unknown

U.S. Department of the Interior Museum, INTR 03001

 The painting at the Interior Museum

c. 1876–81

1876

“smoke or steam rather than clouds”

A possibility in the museum’s teaching guide. Painting image linked at the holder.

George Inness

Saco Ford: Conway Meadows

Canvas catalogued 1876; documented Conway campaign in 1875

A contemporary timetable supports railway service to North Conway. The museum’s possible plume remains unidentified, and regional rain cannot date the canvas’s sky.

Railway context supportedPainted weather unknown

Mount Holyoke College Art Museum, 1883.55.I(b).PI

 The painting at Mount Holyoke

1877

1878

1879

1880

c. 1881–1882

1882

1883

1884

1888–1889

1891

1892

1893–1901

ca. 1896–1897

1897

Claude Monet, The Seine at Giverny (L’Île aux Orties): trees mirrored in still water under a pale pink sky.

Claude Monet

The Seine at Giverny (L’Île aux Orties, Giverny)

Dated 1897; the Île aux Orties at the mouth of the Epte, in the second summer of Monet’s dawn work on the river

The Met says weather held up Monet’s dawn pictures of the Seine in 1896. The records show that first summer ending badly: September 1896 was the wettest in 152 years at Paris and in 37 at Évreux. In 1897 the gauge across the river had its wettest August in 11 years. The dawn mist was never measured.

“Inclement weather” partially agreesDawn mist unknown

Columbia Museum of Art, gift of Mary T. Chambers

 The painting at the Columbia Museum of Art

1899

1907

c. 1907

1908

c. 1909

1911

“Date made: 1911”

PAAM’s catalogue. Image rights unverified.

Margery Austen Ryerson

Commercial Street, Provincetown

Catalogue year 1911; leafy season is Nephos’s reading

Nephos reads a sunlit street in a changeable summer. Blue Hill’s school-season light records support that reading’s seasonal plausibility; local rain and a contemporary gale report reveal another side of the summer.

Painted day unknownRegional light recurs

Provincetown Art Association and Museum, 3282.058.Pa25

 The object at PAAM

1917

1916 / 1917

“it is tempting to speculate that the snowfall of March 11 inspired Roof with Snow; perhaps the sketches were painted that day.”

H. Daniel Peck, American Art 21, no. 3 (2007), p. 106

Georgia O’Keeffe

Roof with Snow

Catalogue year disputed; Peck’s proposed snowfall of 11 March 1917 tested

Amarillo recorded no precipitation and a 73 °F high on 11 March 1917, the day Peck proposed for an all-day snowfall. Its one measurable snowfall after the move fell on 3 March. The painted day remains unknown.

11 March snowfall contradictsSnow after the move partially agrees

Amarillo Museum of Art, CR 126 (catalogue raisonné; holder accession unverified)

 The work in Access O’Keeffe Amarillo Museum of Art

1919

1920

Golden wheat shocks stand across a field before leafy trees and red-roofed village buildings beneath a pale sky.

August F. Biehle Jr.

Wheatfield in Zoar

1920 by the museum catalogue; the depicted harvest stage is Nephos’s reading, not a dated session

Nephos reads cut wheat standing in shocks to dry. Contemporary reports place the regional harvest in July; Dover recorded 8 measurable rain entries totalling 2.97 inches during 7–20 July. The fair-looking scene fits a rain-interrupted harvest, without identifying its day.

Regional harvest rain-interruptedPainted day unknown

Canton Museum of Art, 86.18

 The painting at Canton

About 1920

1921

1927

Mean July–August daily highs at Turbine and Parry Sound across their records, with 1927 highlighted near the cool end. Weather chart, not Lismer’s painting.

Arthur Lismer

Pines, Georgian Bay

Dated 1927, purchased 1928; summer 1927 is the latest summer it could draw on

Dated 1927, but no source found so far places Lismer on Georgian Bay that year. The latest summer the canvas could draw on was cool at both Georgian Bay grounds named in the Lismer sources: 6th coolest of 60 complete summers at Turbine, near McGregor Bay, and 4th coolest of 100 at Parry Sound, near Go Home Bay. The painted summer and bay remain unknown.

Latest summer: 1927Painted summer and bay unknown

University of Saskatchewan Art Collection, 1928.001.001

 The painting at the University of Saskatchewan

1929

“You can see six or seven thunderstorms going on at the same time.”

Marin to Stieglitz, from Taos, July 1929

John Marin

Taos Canyon

Dated “Marin 29”; his first Taos summer, early June to at least 25 August 1929

The rain gauge inside Taos Canyon caught 3.0 mm in June 1929, its driest June; from 8 July the thunderstorms came, and the summer ended the 3rd wettest in its record. Marin wrote that July of six or seven thunderstorms at once. The painting is not reproduced: the Gallery does not offer it openly.

“Six or seven thunderstorms” partially agreesDay unknown

National Gallery of Art, 2005.98.141

 The painting at the National Gallery

1930

c.1930

“Thawing”

Richard F. Oakley, Taos observer sheet, 25 January 1930; snow still measured 2 inches deep.

Walter Ufer

Winter in New Mexico

c.1930; no particular winter established

Taos records at least 37 clear days with snow lying in 1931–32, and 27 in 1928–29. These are strong environmental candidates for Nephos’s reading of the picture, not dates for the painting.

Sun and snow documentedPainted winter unknown

Harwood Museum of Art, University of New Mexico, accession unconfirmed

 The painting at Harwood

c. 1930

1931

June–September precipitation at Portland city across its archive, with 1930 highlighted and incomplete seasons marked. Weather chart, not Graves’s painting.

Abbott Fuller Graves

A Kennebunkport Doorway (Peace and Sunshine)

Catalogued 1931, sold by 7 April 1931; summer 1930 is the latest possible summer

Sold by 7 April 1931, despite its 1931 catalogue date. The latest possible summer, 1930, was relatively dry at Portland: 247.2 mm, ninth driest of 44 complete seasons, with 52 of 122 days classified clear. The actual painted summer remains unknown.

Latest summer: 1930Painted summer unknown

Collection of the Art Fund, Inc. at the Birmingham Museum of Art, AFI.104.2015

c. 1940

1941

1949

“I'd go back home between stages at the sanatorium”

O’Malley, Magill interview, 1984; general recollection, not a date for this picture.

Tony O’Malley

Landscape, Kilkenny Sanatorium, Kells Road

1949; winter is Nephos’s visual reading, not a documented month

Nephos reads winter in the bare trees and grey sky. Kilkenny’s early window was dry; late 1949 brought frequent light rain. A separate Dublin record shows no exceptional frost or sunlessness.

Winter: Nephos’s readingPainting day unknown

Butler Gallery, Kilkenny, object number not confirmed

 The painting at Butler Gallery
“The painting is of the following morning.”

Jennifer Pye, Monhegan Museum of Art & History, email to the author, 13 October 2026. Image at the museum.

Andrew Winter

Wreck at Lobster Cove, Monhegan, Maine

1949 canvas; scene attributed by the museum to 6 November 1948

The museum’s morning-after scene has regional fog context: Portland reported 100–200 m visibility at 06:00–11:00 EST on 6 November 1948. Local weather and Winter’s treatment of it remain unresolved.

Regional fog contextPainted weather unresolved

Monhegan Museum of Art & History

 Monhegan Museum
“The St. Christopher wrecked at Lobster Cove on the evening of August 5, 1949.”

Jennifer Pye, Monhegan Museum of Art & History, email to the author, 13 October 2026. Image at the museum.

Joseph De Martini

The Wreck of the St. Christopher

1949 canvas; wreck dated by museum to 5 August 1949; depicted moment unknown

Portland’s 5 August 1949 evening reports show 16–19.2 km visibility and 1–1.5 m/s wind. This is context for the museum-attributed wreck date; the local weather and painted moment remain unknown.

Event date attributedPainted weather unresolved

Monhegan Museum of Art & History, 5745

 Monhegan Museum

1950

1954