Whilst painting Ophelia, Millais drew straight on to the canvas with pencil to mark out where the figure and large parts of the landscape would be. Close examination of the painting allows us to see his pencil marks at the edge of the painting. By looking under the frame at the spandrels (the curved areas at the top left and top right of the painting, normally covered by the frame) we can learn more about his working practice (see images below).

Left spandrel from Ophelia

Left spandrel from Ophelia

© Tate, London 2003

Right spandrel from Ophelia

Right spandrel from Ophelia

© Tate, London 2003

 Millais is thought to have painted a water vole in his painting on 28 October. He is then thought to have rubbed it out the following day, re-painted it again from 6 to 7 November and eventually removed it on the advice of his friend CR Leslie (a picture hanger for the Royal Academy Exhibition of 1852).

The sketch for the water vole under the right spandrel may be a study of this creature that he decided not to include. The creature on the left spandrel may be a study for the water vole too, or perhaps a squirrel? Millais did admit to having trouble painting the rat. Hunt’s uncle and aunt mistook it firstly for a hare, then a rabbit, then a cat or a dog!

Also on the left spandrel we can see two sketches of birds, perhaps studies for the robin that he included among the branches of the willow tree. There are some trials of paint and some writing that reads, ‘55 Chancing Lane’, perhaps a house of a friend?

And on the right spandrel, we can see another sketch for a bird, and some writing that reads ‘loosestrife’, which is the purple flower in the top right of the painting.