I'm a huge fan of Richard Dawkins. Despite the claims of his detractors, he is consistently calm and polite when arguing with people who disagree with his views, and his books -- oh, if only his detractors would read and understand them! -- are all lucid, thought-provoking and educational. For Dawkins to produce a book aimed at instilling in young readers a sense of wonder in the magic of the real world was a bold but commendable step. His approach, outlining the myths used by superstitious people to explain what they don't understand then showing how the real explanations are both more satisfying and convincing, is original and effective.
The problem with the book is that it only sometimes achieves what its cover says it intends: to explain HOW we know what's really true. Dawkins has run up against the obstacle that confronts every science teacher at every level. Science has given us so huge and so deep an understanding of our planet and the universe that it is by now impossible to detail the evidence for everything we know to be true. The consequence is science teaching that is often decried as a "wall of facts". There is so much to be learned it allows little room for presentation of the people who made the discoveries and the evidence on which the discoveries were based.
Newton's laws of gravity, Darwin's theory of evolution and Einstein's theories of relativity retain the names of the people who assembled the evidence, but for most familiar scientific "facts" we no longer have any idea whose work and what evidence lay behind their discovery. It is therefore disappointing that a book that sets out to explain how we know what is real so often follows the wall of facts approach. Chemical elements and compounds are described without even a hint how we proved the difference between the two. Crystal structures, behaviour of molecules in solids, liquids and gases, subatomic structures and the bonding flexibility of carbon atoms are all described (beautifully) but not accounted for with evidence. A Dawkins phobic reader would be entitled, for many chapters, to say "so we have to believe this just because it's written down in your book?" Which I think counts as an own goal.
When the book does get into scientific evidence it does so with finesse. The chapter "What is a rainbow" beautifully explains how Newton showed white light is made up of the spectral range of colours. The ingenuity of Newton's work with light beams and prisms leaps off the page at you. And the chapter sets the stage for understanding subsequent accounts of stars and galaxies. If only the same approach could have been used for every part of the subjects covered: but then it would have become a giant book.
Dave McKean's illustrations are brilliant, and the book's layout is so carefully organized that the text amounts to a flowing set of figure legends. Thus the one occasion when a separate caption is given for a figure jars the reader. The incongruous caption in question appears in the chapter on immunity and is made worse by containing an error. The illustration clearly shows antibodies binding to a virus surface, while the legend states that immune T-cells have attached themselves to the virus. This howler is surprising, considering the many colleagues Dawkins could easily have consulted for a cross-check.
For people who really want to read about how evidence for science was obtained, Bill Bryson's "A Short History of Nearly Everything" is an excellent, though not outstanding, attempt. Or you could read Simon Singh's masterpiece "The Big Bang" and get a real sense of the intellectual battle that rages within any scientific arena as new evidence constantly advances the ability of Homo sapiens to comprehend reality.