Under the spells of scientific discovery

Publishers have realised that science sells and that anthologies do too. John Carey's Faber Book of Science was a best-seller, and this me-too from Little, Brown also deserves to do well.

As might be expected, this volume has many treasures. The last two pieces in it are Richard Preston on the asteroid-hunters Gene and Carolyn Shoemaker from Preston's First Light , and the classic chapter on carbon from Primo Levi's The Periodic Table (also in Carey's collection).

Both of these are terrific pieces of writing by any standard, but the book also contains more than its share of material that is at best second-rate. For example the first piece, on chemistry by Isaac Asimov, shows clearly why Asimov's reputation is fading.

But despite the presence of some clunkers, editor Edmund Bolles has worked hard. His passion appears to be early rather than modern science writing. Galileo, as well as providing the title (in fact from Brecht's play about him, not Galileo's own words) gets three pieces. One is his exciting account of the first telescope views of the sky, which still reads well. And, mysteriously, two pieces, separated by over 200 pages, are from his famous cod debates on the new science between the wise and foolish men. By contrast these have aged badly and seem very mannered.

Despite a baffling attempt by Bolles to explain the pattern of the book, the 60 pieces appear in essentially random order, so that astronomy may be followed by behavioural science or quantum theory, and a practical piece by a theoretical one. This is the right approach, since the pleasure of reading an anthology such as this is to a large degree in the element of surprise.

Along the way the reader is struck forcibly by the different staying power of the writing on display. Herodotus on the creation of Egypt by the Nile (444 BC) has aged a lot better than Gordon Allport (AD 1937) on the "mature personality", a piece that now seems so subjective and data-free that an editor ought to have left it out or explained its inclusion better than done here.

By contrast, Johannes Kepler's 1610 writings about the moon are little read today but are perhaps the founding document of what is now a substantial science, planetary geology. He compares the lunar terrain with the valleys of Austria much as our contemporaries might look directly at the similarities between the volcanoes of Mars and of the Earth. The same applies to Leonardo, whose realisation about the significance of sea-shell fossils in high mountains has still not sunk in with the creationists of our own era. By contrast, Harper ' s magazine's 1902 piece on radioactivity is written in an ornate style that it is hard to believe was in fashion during the present century: "But men are children, and one question fathers another," and much more in this vein.

Radioactivity is also represented by Marie Curie's memorable 1923 article on the isolation of radium. This is one of a regrettably small number of pieces in the book on how scientists actually do science, and in this case overcome spectacular obstacles on the way, including working in an abandoned building only good for demolition. Sometimes she and her husband would visit the lab at night and watch the cheery green glow of the radioactive bottles and jars around them. Only later did anyone think that radiation and people are not meant to mix. In the same way, we all know that Pavlov kept dogs waiting for their dinner, but it turns out in his own account that his dog experiments were cruel and not particularly informative by our standards.

Bolles has done his best to combine writing by scientists and by writers about science, and the lesson is that good science writing can come from either.

Another lesson from the book is that scientific thinking, rigorously applied, can illuminate even subjects about which there is next to no primary data. For proof, see Carl Sagan's look at UFOs. Here he uses arguments analogous to the proof that there is no Santa Claus. Even if Santa could visit one home per second it would take more time than he has to visit every child who gets presents. You do not need to know about reindeer biology or elf evolution to prove the point. In the same way, it can be shown that any realistic project to send flying saucers around the galaxy would need the materials from millions of stars, a piece of engineering that seems not to be occurring. You need no knowledge of alien cultures to know that this more or less disproves the UFO hypothesis, especially as it is allied to the complete lack of useful or interesting recovered UFO technology.

Bolles points out that few people realise what hard work science is. But in an excerpt from The Double Helix , James Watson observes that key influences can come from anywhere. His collaborator, Francis Crick, went to a lecture by astronomer Tom Gold in which Gold talked about the perfect cosmological principle. Musing on this, Crick developed the idea of a perfect biological principle that would allow the self-replication of genes. Without this inspiration, the structure of DNA would still have been found - but not so soon, and perhaps someone else would have got the Nobel prize.

Galileo's Commandment is full of good things (the earth and space sciences, and chemistry are especially well represented) but it loses out to Carey's anthology on the grounds of its breadth of coverage and even more so for its chaotic organisation and for the editor's verbose and often unnecessary interventions. This is a field where more is less: if the pieces are chosen properly they should need only a minimum of explanation.

Martin Ince is deputy editor, The THES .