Purkinje images

How it works

Light entering the eye crosses four curved surfaces: the front of the cornea, the back of the cornea, the front of the lens, and the back of the lens. Each surface reflects a small amount of light back out. So a single bright light held near the eye produces four separate reflections. These are the Purkinje images, numbered one to four from front to back¹.

The first image, from the front of the cornea, is the bright dot you see on someone’s eye in a photograph. It is by far the strongest. The second and third are dim and hard to see without instruments. The fourth image, from the back of the lens, is special. The back of the lens curves the other way, so it acts like a concave mirror and the fourth image is upside down¹.

The images move in different ways when the eye rotates, because the surfaces have different centres of curvature. Video eye trackers use the first image together with the centre of the pupil to work out where a person is looking². Older laboratory trackers used the first and fourth images together, which cancels out small head movements².

The third image changes size when the lens changes shape to focus on a near object. In the 19th century this was the evidence that the lens really does change shape during focusing¹.

There is also a fifth reflection that only the owner of the eye can see. Light bounces off the retina, reflects off the back of the cornea, and lands on the retina again. In a dark room, with a small bright light and a relaxed eye, some people see it as a faint second copy of the light¹.

History and culture

Jan Evangelista Purkyně described the reflections in 1823¹. Louis Joseph Sanson used them in the 1830s to study the lens, which is why they are also called Purkinje-Sanson images¹³.