On Being, by Peter Atkins (Oxford, £10.99)

Do you think all possible questions about death are answered by an explanation of how a corpse decomposes? Then this brief catechism of rampant scientism is for you. The chemist-author sets out boldly to show that science can explain everything and dispel all mysteries: it can even "elucidate love, hope and charity", Atkins claims, without showing how. Instead he discusses cosmology, evolution, sexual reproduction, and the eventual extinction of life on our planet, delivering with satisfaction the anti-consolation that "we [. . .] are inescapably destined to decay", a claim that itself oversteps the bounds of dispassionate scientific description.

Atkins wants to celebrate the sophistication of current scientific understanding (there is a terrific description of "what went on at a molecular level when I blinked"), but also to advertise the "near-spiritual" beauty of anti-spiritualism, though the going is not quite so near-spiritual during his snide and pointless tirades against religion. One moment of nostalgia strikes a picturesque note: for the author, "a dead frog pinned to a dissecting board" symbolises a bygone age when "real science [. . .] was studied" in schools. I seem to sense the nervous swivelling of amphibian eyes.

Moon: A Brief History, by Bernd Brunner (Yale, £18.99)

Philosophy and myth are not so dispensable for this author: you can't, after all, explain everything about the Moon without considering what it has meant to human beings over thousands of years. Brunner's perky cultural history  of the Moon in superstition, song, and indeed science  encompasses many wonderful things both imaginary (inhabitants including man-bats or cat-women) and actual (strange lights known as "lunar transient phenomena", or the burial on the Moon of an American geologist's ashes).

Brunner illuminates current theories of the Moon's formation as well as long-held fantasies of colonising it, including a secret US plan in 1958 to nuke the Moon as a show of power. We see Galileo exaggerating reality in his maps, becoming a "space artist" avant la lettre; and enjoy a brief history of the centuries-long competition to name lunar features. I was particularly taken by the tale of an Italian writer who proposed, in the 16th century, that writing could be projected on to the Moon using a giant mirror. Happily it was unfeasible, or the Moon would now be covered in adverts.

Twelve Examples of Illusion, by Jan Westerhoff (Oxford, £18.99)

By the way, why does the Moon look bigger when it's near the horizon? The answer is not simple, as we learn in one chapter of this intriguingly eclectic book. Starting with a Buddhist list of 12 types of illusion, Westerhoff mixes Tibetan and Indian parables with European philosophy and findings from modern psychology on unreliable perceptions. The pages are populated with lovely colour pictures and stylish paradoxes; lucid dreams, mirages, echoes, bubbles, mirror images and more are explored with a pleasantly provocative insistence on multiplying difficulties.

Oddly, Westerhoff at one point describes atoms as "theoretical entities", in contrast to "untheoretical ones, like tables and chairs". Perhaps all he means is that atoms cannot be seen with the naked eye. (Yet the naked eye guarantees nothing, insists the whole book.) Perhaps he does not believe that tables are made only of atoms. (For if they are, and an atom is theoretical, then a table made only of atoms can be no more than theoretical.) Or perhaps "theoretical" is a term of praise (it's a pretty good theory, after all), and the problem is really that we don't have a good theory of furniture.