A little cocaine for the eyeball helped explain those jerks in our sight

When the history of science became a distinct profession in the middle of the 20th century and university departments were created, the practitioners were still concerned primarily with the history of science itself, the tracing of the sequence of scientific understanding. Nowadays, however, many historians of science - there are distinguished exceptions - are concerned with the social milieu of science, with the changing status of the natural philosopher and the role of institutions and networks. And so once again the history of particular sciences, the history of the conceptual unfolding of particular disciplines, is often written by active scientists.

This is certainly true for the history of sensory theory, the history of our understanding of visual and auditory perception. In part, this is because some historical experiments in this field are of continuing validity and so the modern scientist becomes aware of the field's history through the normal scientific etiquette of citing the origins of ideas and facts. Nicholas Wade of Dundee University is characteristic of the breed of scientist-historian: he is respected for his own experimental work on visual perception, but he has also established himself as one of the foremost historians of visual science. His latest books are historical: Perception and Illusion is an overview of the history of perceptual theory; The Moving Tablet of the Eye (written with Benjamin Tatler) describes how eye movements came to be measured and understood.

Perception and Illusion begins in the time of Euclid and Aristotle, when it was still uncertain whether visual perception depended on something emitted from the eye or on something transmitted from the object to the perceiver.

Wade carries us forwards via the remarkable Arab writer Alhazen to the understanding in the 17th century of image formation and optics, and the introduction in the 19th century of laboratory instruments for studying the senses. He ends with the late 20th-century view of the human observer as an information-processing system able to construct internal models from a dynamic binocular input.

Despite this vast territory, Wade has a firm and accurate grasp of his material. One error, however, is in his account of Goethe's first experiment with a prism. The poet had borrowed a prism from his friend Christian Wilhelm Büttner but had neglected to make any observations with it. "When asked to return the prism," Wade writes, "he simply directed it to a light and concluded that it still looked white." Now, if Goethe had looked at a point of light through the prism, he would have seen an extended spectrum. What he did, by his own account, was look through the prism at a white wall: the extended white surface continued to appear white (as it should by a Newtonian account) and colour was apparent only at edges, being liveliest on the cross frames of a window. Goethe was led at once to suppose that the Newtonian theory was mistaken and that colour arose, as ancient writers had held, from the interaction of light and dark.

There is just one very odd thing about Perception and Illusion : there are no pictures. Although the author is known for his graphic skills, this book on the history of perception contains not a single anatomical diagram, portrait or illusion. This must have been a conscious decision.

Perhaps it is to remind us that there are no pictures within our visual brain. But we must hope that one day Wade will produce a fuller version of this book that is illustrated in his usual style and less hurried in the writing.

The Moving Tablet of the Eye is more handsomely produced and illustrated, and it comes at a timely moment. Eye-movement recording was traditionally a fiddly business, requiring constant attention to alignment and calibration.

But the recent development of infra-red eye-trackers, often using video imaging of reflections from the eyes' surfaces, has brought an explosion of eye-movement recording, not only for the purpose of analysing our most exquisite motor system for its own sake, but more generally for understanding the cognitive and social processes revealed by the changes in our gaze. Eye movements are reported to be abnormal, for example, in those affected by schizophrenia and Asperger's syndrome. Those who carry out eye-movement research, and who would like to learn more about their forerunners, should read The Moving Tablet of the Eye .

Introspection proves a poor guide to the nature of our eye movements. We may have the subjective impression that our eyes move smoothly around a scene or along a line of text, but in fact they move in sharp jerks, or saccades, between successive points of fixation. And it was only at the end of the 19th century that Benno Erdmann and Raymond Dodge observed that one can never catch one's own eye moving in a mirror: during each saccade, one is in effect blind.

So objective methods of recording eye movements are necessary, and Wade and Tatler survey the varied methods that have been used. The most heroic must be that introduced by Edmund Delabarre in 1898. Realising that plaster of Paris will attach itself firmly and immovably to any moist surface, Delabarre made a plaster cast of the cornea of an artificial eye, formed a central aperture corresponding to the pupil, mounted the cast on his subject's eye, and linked it mechanically to a kymograph. It was necessary to render the eyeball anaesthetised with cocaine; and Delabarre writes:

"The plaster will not detach itself until it becomes thoroughly soaked with tears." It was necessary to allow a week between experiments.

A gentler method of rendering eye movements visible is to form an after-image and then subjectively monitor its movements relative to objects in a scene: the after-image is fixed on the retina whereas the images of real objects will move across the retinal surface.

This technique was used by the American-born William Charles Wells, whom Wade and Tatler identify as a founder of eye-movement research. Wells, an "ingenious but irritable" figure, was known for his anticipation of Darwin's theory of natural selection, but Wade and Tatler have rescued from neglect his pioneering study of the nystagmus that follows rapid rotation of the body. This is a repetitive eye movement - slow in one direction, saccadic in the other - of which the vertiginous subject may be quite unaware.

The saccadic nature of our normal eye movements, around a scene or along a line of text, was discovered only towards the end of the 19th century. Wade and Tatler spend some time challenging the credit traditionally given to Emile Javal for this discovery; they give more weight to the contributions of Jarval's colleague Lamare and those of Ewald Hering, Alexander Crum Brown and Dodge.

Among the vices of the scientist-historian, the historian of science would certainly list an unwholesome interest in priority. Scientist-historians are too much concerned with who first proposed the correct theory, who first designed the critical experiment. But that is natural for us. Science is a competitive sport.

John Mollon is professor of visual science, Cambridge University.