Castle Geyser, Yellowstone

Step 1 · How the picture was made

A dated sketch, a later canvas

The Eiteljorg Museum’s painting is an oil on canvas dated 1891. The museum’s account of its making gives a more precise field date: Brown sketched Castle Geyser on site on 6 September 1890, then completed the canvas in his Helena, Montana, studio in 1891. The dates describe successive stages of work. The weather comparison belongs to the sketch date, not automatically to the sky in the finished painting.

Brown advertised Yellowstone views as “careful studies from nature”, according to the MFA Boston’s transcription of his 1886 brochure. That catalogue concerns a different painting and documents how he presented his Yellowstone work. The Eiteljorg’s object-specific history is the basis of this study.

Step 2 · The observation on that date

Snow at Mammoth Hot Springs

On the row dated 6 September 1890, the Mammoth register records a maximum of 45 °F (7.2 °C) and a minimum of 28 °F (−2.2 °C). The printed superintendent report also records snow covering the ground. The manuscript Army sheet agrees on both temperatures and the snow remark.

This is Mammoth Hot Springs, not Castle Geyser in the Upper Geyser Basin. The manuscript is headed Camp Sheridan; the following year’s printed report labels the series Fort Yellowstone. These are two versions of the same station record, not two independent weather witnesses. The thermometer reset hour is not established, so the extrema remain attached to the source’s date rather than a reconstructed painting session.

Step 3 · The month around the sketch

The lowest daily high in this month’s register

The sketch-date maximum is the lowest of all 30 daily maxima in Mammoth’s September register. This comparison places the date within that one month. It is not a ranking against other years, a modern climate anomaly, or a measurement of the temperature at the geyser.

Step 4 · Myth and reality

A real place does not make a canvas a daily log

The documented fieldwork connects this studio painting to an encounter with Castle Geyser. The weather records add a specific fact: snow was recorded elsewhere in the park on the sketch date. They do not establish snow at Brown’s easel, or show that he removed it. No local weather observation or sketch hour has been verified, and the field sketch has not been inspected. We cannot identify which features of the scene he retained, combined or changed in the studio.

Nor can the plume be used here as a thermometer or wind gauge. No local humidity, wind-speed reading or dated eruption log has been connected to the sketch. The National Park Service notes that Castle’s eruption pattern has changed over its recorded history; today’s eruption intervals cannot date an 1890 scene. No painting-specific scholarly description found in this research establishes a cloud type, light condition or weather event for this canvas.