Kaaterskill Falls

Step 1 · The picture’s history

First encounter, return, studio

Cole first travelled up the Hudson in 1825. His first painting of Kaaterskill Falls, bought by John Trumbull, is now lost. Noble’s 1853 biography places the journey in autumn; the Met’s 2018 chronology places it in summer. The bookstore display that October–November and Trumbull’s letter of 14 November 1825 put the first canvas before that date; they do not settle the season of the trip.

The Wadsworth picture belongs to the following year. On 6 July 1826, Cole wrote to Daniel Wadsworth of sketching around Lake George and the Catskills during the preceding month or two. The Met chronology connects those new sketches to this canvas, its figure 146, accession 1848.15. In November, Cole told Wadsworth how difficult it was to repeat a view. The painting was made in the studio; no painted day is documented.

The foliage is painted in autumn colours. Gaudio calls this an “autumnal landscape view” (p. 457). That autumn does not match the May–early July 1826 return window. It is consistent with Noble’s autumn account of the first trip in 1825, or with a pictorial choice; the records cannot decide between them. Gaudio also describes a “clearing sky” in the distance (p. 453). These descriptions concern the painted scene. The weather window covers Cole’s return, not the season or weather depicted.

We use 6 May–6 July 1826, 62 days inclusive, as an approximate reading of the July letter. It is a window for the return journey, including Lake George, not an itinerary at the falls. No 1825 weather figures are shown: Watervliet begins on 1 August, and its September readings need further checking.

Step 2 · The water question

A dry May, then a wet June—at Albany

The fall pours past the cavern opening and breaks into mist. Michael Gaudio’s account of this painting places us behind the upper fall, looking out towards the mountains. Was the spring and early summer of Cole’s return wet or dry?

Albany Academy’s register, kept by T. Romeyn Beck and published by Franklin B. Hough, gives 0.76 inches in May 1826, the driest May of its 24 years, 1826–1849. June has 6.22 inches, fourth wettest among the same 24 Junes. The Hudson-valley record changes direction between the two months; it does not support calling the whole return uniformly wet or dry.

These are whole calendar months. May includes five days before our approximate window; June falls wholly within it. We have no verified rainfall total for the first six days of July, or for the falls’ catchment. The monthly figures cannot tell us when rain fell during Cole’s travels.

Step 3 · Context for the return

A warm start to summer

At Watervliet Arsenal, 65.4 km from the falls in a straight line, the thermometer was read at 7 a.m., 2 p.m. and 9 p.m. All 186 readings survive for the 62-day window in 1826. Their mean is 69.5 °F, second warmest of 12 complete, unflagged windows in the 1826–1840 period with that same schedule. This describes warmth in the region during the return; temperature does not answer the water question.

The comparison years are 1826, 1827, 1828, 1830, 1831, 1833, 1835, 1836, 1837, 1838, 1839, 1840. Windows in 1829, 1832 and 1834 have missing or flagged readings and are excluded. These are fixed-hour means, not means of daily highs and lows. Watervliet’s rain cells are blank in this window; blank does not mean dry.

Step 4 · Where the records stop

The water in the canvas remains undated

The records give the return journey a regional setting: a dry May and wetter June at Albany, with warmth at Watervliet. They cannot connect the painted torrent to a particular rainfall event. Neither the date of Cole’s renewed visit to the falls nor a local discharge measurement is established. A studio picture that repeats an earlier view can preserve an encounter without preserving the weather of a single day.