Both of these books fall foul of the Trades Description Act: they purport to examine the light science throws on art and vice versa, but there is little mention of art in either. They leave the dichotomy between art and science as sharp as ever.

Science seeks rigorous theories that enable anyone who has grasped them to draw the same (correct) predictions. Elegance, simplicity, symmetry and breadth are also among the hallmarks of a good scientific theory. None of these is necessary in art. Though decorative art is usually elegant and symmetrical, many paintings are not: indeed they may attempt to move us by their very ugliness - compare the work of Francis Bacon or Goya. Nobody can specify what constitutes harmony in a painting nor why some paintings move us and others do not. In representational art, it is easy to point out parts that are badly drawn, but - as in El Greco or Picasso - the distortions may be deliberate to create an effect. As a matter of fact, the influence of science on art has been much greater, at least in recent years, than that of art on science. The Bauhaus school learned from the Gestalt psychologists' laws of grouping, Escher was inspired by the concept of figure and ground, and architecture has been radically changed - often for the worse - by technological advances in materials. None of this is the concern of either The Artful Eye or The Artful Universe.

Richard Gregory in his "pretext" to The Artful Eye claims the book will "cast the eye of science on art" and also that "the scientist has at least as much to learn from the artist as the other way round". In fact only two of the 21 contributions are about art - one on Vermeer and one on forgeries.

Philip Steadman attempts to show that the perfect linear perspective of Vermeer's paintings of his studio was achieved by the use of a camera obscura, a lens that images the scene on a wall or screen, from which the artist can make a perfect copy of the scene as observed from the position of the lens. The author's reconstruction of a three-dimensional model of the room is ingenious, but hardly epoch making, for it merely confirms a widely held belief. As to forgeries, they are nowadays most readily detectable by scientific methods. David Phillips argues that art critics are taken in because they approach the picture with high expectations: if it is genuine they have at least made an important find and at best a lot of money. If some forgeries are not distinguishable by eye from the work of the original painter, one wonders why they are not more highly valued.

ÁñÁ«ÊÓƵ

The Artful Eye is devoted almost exclusively to developments in visual science, with here and there a nod in the direction of art. For example, there are cells in a monkey's cortex that respond specifically to faces. To make this finding relevant to art, one contributor suggests it is because these cells exist that so much art depicts faces: however, he immediately withdraws, pointing out correctly that the face detectors and the frequency of faces in paintings may be determined independently by the importance of faces for social animals such as monkeys and men. In fact, the chapter by David Perrett and others on the properties of these cells is one of the most interesting in the book, for they closely mimic our own capacity to recognise a face as a face. For example, the cells do not respond as much to inverted faces as to faces the right way up and they respond only weakly to negatives of a face (in which dark and light areas are reversed) and to faces lit from below rather than from above. These are exactly the conditions under which people have difficulty in detecting the presence of a face. Moreover, like us, the cells even respond to line drawings of faces. All this is very interesting, but knowing where in the brain these cells are and knowing that they detect faces in much the same way as we do does not solve the most important problem: how are they wired up to do this? In fact, despite all the psychological and physiological advances on vision since the 1950s, we know little more now about how objects are recognised than we did then.

There is a good chapter on the use of shading to extract the three-dimensional structure of an object. A circle that is light on top shading to darkness at the bottom is likely to be an image of a sphere since light normally comes from above. If the shading is reversed it is likely to be a hollow. But if the light comes from either side, it will be ambiguous (either sphere or hollow depending on which direction the light is coming from). V. S. Ramachandran shows that people make three implicit assumptions in interpreting such patterns. They assume, first, that objects are normally lit from on top; second, that light comes from a single direction - where there is a whole series of circles some bright to the left others bright to the right, the set of circles is seen in a way consistent with the assumption that light comes from one direction or the other; third, that where there is complete ambiguity, we interpret circles as bumps not hollows. It should go without saying that artists have made extensive use of shading as a depth cue, but of course the author cannot resist making this point.

ÁñÁ«ÊÓƵ

It has been known for some time that different features of the visual image are extracted in different parts of the visual cortex, including the orientation of lines and edges, colour, stereopsis and motion. In several chapters this idea is carried even further - different aspects of what is apparently the same feature are analysed in different cortical areas. Some examples are broad versus narrow stripes and fast versus slow motion. Moreover, Margaret Livingstone and David Hubel, in an exceptionally lucid chapter, argue that what appear to be complex high-level features, such as linear perspective and illusory contours (contours created where there is no difference in brightness) are extracted at a low level of the visual system. Both effects disappear if the stimuli differ from the background only in colour but not in brightness: this suggests that these two features are extracted before colour has been analysed.

Despite the interest of much of the work reported, the book has faults. It needs severe editing to prevent repetition and to clarify many of the chapters. It is highly technical and will be unintelligible to the general reader, though it sometimes gives information that is insulting to the specialist. Some passages are unintelligible to anyone: for example, Colwyn Trevarthen writes: "Art comes from and appeals to vitality in awareness and it is evaluated by innate 'vitality effects'. It has traces in it of the spaces and time of all behaviour and the form of the body in action, with all its complementary senses". Well, fancy that. Moreover, each chapter is highly specialised: the reader is never given a general view. One of the most pressing problems in visual physiology is not even mentioned: given that different features are extracted in different cortical areas, how are they put together to form a composite interpretation of an object? With the possible exception of colour vision, the more we know about the neurophysiology of vision, the more intractable the problem becomes.

John Barrow's The Artful Universe claims "to explore the close ties between our aesthetic appreciation and the basic nature of the Universe", but again it gives little space to art. Its main topic is the way the universe as it exists both on a cosmic and planetary scale has shaped mankind physically and mentally. Although Barrow hops, seemingly at random, from one topic to another, much of his material is fascinating. We could only exist in a vast universe because it takes the stars billions of years to produce carbon: hence to support life (at least as we know it) the universe must have been expanding for billions of years and therefore must be very large. On an intermediate scale were the earth not to have the drag of its moon, it would be rotating four times as rapidly.

The result would be intense winds causing severe erosion, preventing the growth of all but the smallest trees and causing plants to attach their leaves much more firmly. The effects on the Earth's ecology are not wholly predictable, but the conversion by plants of carbon dioxide into the oxygen we need might have been long delayed. Again, with no moon there would be no tides and the ecology of the seashore from which man's distant ancestors emerged would be very different. Barrow might have added that without tides, offshore pollution would now be so great as to be insupportable.

On a smaller scale yet, he points out that the size of animals is limited by the force of gravity. Volume (and also mass, assuming constant density) is proportional to the cube of length, while area is proportional to its square. Hence, to support a very large mass, legs would have to be impossibly thick. These ideas, and many others, are both entertaining and thought-provoking.

ÁñÁ«ÊÓƵ

When, however, he leaves physics, his views become more questionable. In colour vision there are two "opponent processes" - blue versus yellow and red versus green. He ascribes the evolution of the first to two facts - blue light is reflected from the sky and sunlight is strongest in the yellow region of the spectrum. He attempts to explain the red-green opponency by an argument too convoluted to enter into. Unfortunately for his explanation, other animals, including reptiles, amphibia, fishes and butterflies, have colour receptors sensitive to wavelengths very different from our own. Extensive work has found no correlation between the different wavelengths to which different species are maximally sensitive and the environment in which they live.

We must chivalrously assume that Barrow sometimes says what he does not mean, as in his claim that "language is not a learned behaviour". Moreover, his attempt to trace pictorial art back to ancient instincts is highly speculative. He argues that we find flowers pretty because they tell us where other things are growing, but on this account surely we should find a clump of ancestral Brussels sprouts even more beautiful. He may be right in supposing that horror movies fulfil our instinct to explore a dangerous environment from the safety of a chair, but if art is guided by what was important to us in the distant past, why are there not more paintings of log fires, not to mention caves?

The art form Barrow deals with at greatest length is music. Although most of this chapter is conventional, he introduces the notion of a grammar both for tonality and the rhythm of classical music, relying on the ideas of Christopher Longuet-Higgins.

ÁñÁ«ÊÓƵ

There are many topics one might have expected to appear in these two books. There is nothing on the debunking of the golden section by psychologists, on colour preferences or - more tricky - on the clash between certain colours. Even more surprisingly, John Willats's seminal account of the drawing systems used in paintings in different epochs and by different artists is not so much as mentioned. Willats describes the types of optical projection employed by the ancient Greeks, the Chinese, the classical western artists, Cezanne, Matisse and many others, thus providing almost the only clear example of using science to understand what is on the canvas.

Perhaps one should not condemn these two books too severely for their failure to confront art with science. The most important aspects of painting are ineffable and "whereof one cannot speak, thereof one should be silent".

Stuart Sutherland is emeritus professor of experimental psychology, University of Sussex.